Home

Administration Guide - Oracle Documentation

Contents

1. TABLE NM SEQUENCE NAME CURRENT VALUE MIN VALUE MAX VALUE KDD ATTR ATTR ID SEQUENCE 10000000 10000000 19999999 KDD AUGMENTATION AGMNT INSTN ID SEQ 10000000 10000000 19999999 KDD DATASET DATASET ID SEQUENCE 10000000 10000000 19999999 KDD JOB JOB ID SEQ 10000000 10000000 19999999 KDD LINK LINK ID SEQ 10000000 10000000 19999999 KDD LINK ANLYS NTWRK DEF NTWRK DEFN ID SEQ 10000000 10000000 19999999 N KDD LINK ANLYS TYPE CD TYPE ID SEQ 10000000 10000000 19999999 KDD LINK SUMMARY LINK SUMMARY ID SEQ 10000000 10000000 19999999 KDD LINK TYPE SUMMARY LINK TYPE SUMMARY ID 10000000 10000000 19999999 SEQ KDD NODE NODE ID SEQ 10000000 10000000 19999999 KDD NTWRK NTWRK ID SEQ 10000000 10000000 19999999 KDD PARAM SET PARAM SET ID SEQ 10000000 10000000 19999999 KDD PTTRN PTTRN ID SEQ 10000000 10000000 19999999 KDD RULE RULE ID SEQ 10000000 10000000 19999999 KDD SCNRO SCNRO ID SEQ 10000000 10000000 19999999 KDD SCORE SCORE ID SEQ 10000000 10000000 19999999 KDD SCORE HIST SCORE HIST SEQ ID SEQ 10000000 10000000 19999999 KDD TSHLD TSHLD ID SEQ 10000000 10000000 19999999 KDD TSHLD HIST HIST SEQ ID SEQ 10000000 10000000 19999999 KDD TSHLD SET TSHLD SET ID SEQ 10000000 10000000 19999999 284 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Scenario Migration Utility Chap
2. Workflow Number Workflow Name Description 0431 w p0431 build MASD from MASMS This workflow updates the Managed Account Summary Daily Table from its corresponding staging table 0440 w p0440 update nmsm for daily acti The workflow updates Investment Manager Summary vity Month with Daily Data 0450 w p0450 update rrsm for daily acti This workflow updates Registered Representative Sum vity mary Month table from Registered Representative Summary Temp 0460 w p0460 update masm for daily acti This workflow updates the Managed Account Summary vity Month Table from its corresponding staging table 0471 w p0471 update asm for profit and This workflow update the profit and loss related fields in loss the account summary month table 2010 w p2010 update nm from ma This workflow aggregates managed account net worth and count of sub accounts grouping by Investment Manager It updates the Investment Manager table MLM Banking Common Folder 0081 w ph0081 truncate cbsm This workflow truncates Client Bank Summary 0480 w ph0480 aggregate asm to cbsm This workflow performs daily re aggregation of the Cor respondent Bank Summary Month table out of the account summary month stage 2090 w ph2090 update jurisdiction in cb This workflow updates the jurisdiction for CLIENT BANK Correspondent Bank 3130 w ph3130 create client banks from This workflow maintains the Correspondent Bank tab
3. 299 Messape Template Repository sss issssseieisiesnssvesacssassoasssocnsatsesdeaniovedesnaniotenstasesasesasvsedesasoddndedobestssaseiesssosesevensensasenedesions 300 Log unrdboljr 300 Lopoiie Message librae Spiane meo Rae entescedestes tende teuer dede pe ite rater toits 301 Administration Tools issena eerie p iiis REA eR EE HR ROME te ERE RR ERR XH Rope dee Re vasa o ER No e 301 Database M 301 Scenario Manas er 5 Lecce a e EE tiki cea itucehnbnhoitud Surtout s 301 SOLVIGES M 301 Alert Manaeement Case Ma agement us eseuae innin cot is ons ce e putedpre tulit ett dendi ddp onep ed et mat eas utin 302 breue ET 302 Application Server lOOS M A 302 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide xiii Contents Database objects logs A 302 Inpestiom Manas eNi sues udesetebuu eU UD o Sese NEO E AG 302 Lopeine Gonfiguration Ple soe re rt tta dee ee erento te si tote Ge uet edet RARE 303 sample Configuratior Elle coercere tpe nime ee e een e UR EUR RR A 304 Logping Location Property Values riri rtt YR Pe REX REESE OR a E o XE Re Roe beide 307 Log File Sizes P 307 Configurable I ogeing Properties sansen naaie pE ESE IEE a
4. Sequence Number Folder Workflow Number Workflow Name Predecessor 0 ORION_Production 0010 w_p0010_dump_asm_to_file 0 ORION_Production 0070 w_p0070_reload_asm_from_dump_file 0 MLM Brokerage P 2020 w p2020 delete nonstanding instructions roduction Informatica Miscellaneous Pre Watch List Workflows Fraud Detection Pre Watch List Miscellaneous workflows perform processing in support of workflows in multiple functional areas Figure 45 illustrates Pre Watch List Miscellaneous Informatica workflows for Fraud Detection Sequence 1 2005 2140 Sequence 2 Sequence 3 Figure 45 Pre Watch List Workflows Fraud Detection Note Processing can occur simultaneously within each sequence Workflows with predecessors must wait for predecessor results before workflow processing can begin The application completes these workflows in three sequences processing of the workflows does not require data from external predecessors Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 159 Fraud Detection Informatica Workflows Chapter 5 Informatica Workflows Table 62 describes the Informatica Miscellaneous Pre Watch List workflows in Figure 45 Table 62 Pre Watch List Workflows Fraud Detection Sequence Number Folder Workflow Number Workflow Name Predecessor 1 ORION Common 2005 w p
5. Workflow Number Workflow Name Description BSM_Production Folder 0040 w p0040 dump masm to file This workflow is used for the nightly dump of Managed Account Summary month table for the current month to a flat file 0050 w p0050 dump nmsm to file This workflow is used for the nightly dump of Invest ment Manager Summary Month table for the current month to a flat file 0060 w p0060 dump rrsm to file This workflow is used for the nightly dump of Registered Representative Summary Month table for the current month to a flat file 0110 w p0110 reload masm from dump file This workflow mapping rebuilds Managed Account Summary Month from Dump File 0120 w p0120 reload nmsm from dump file This workflow can be used to rebuild Investment Man ager Summary Month from dump file 0130 w p0130 reload rrsm from dump file This workflow can be used to rebuild Registered Repre sentative Summary Month Table from dump file 0171 w p0171 truncate masms This workflow truncates managed account summary daily 0360 w p0360 aggregate block allcn day This workflow is used for the daily aggregation of the trades to file block allocation day trades data This populates the managed account summary file 0370 w p0370 aggregate block allcn trad This workflow is used for the daily aggregation of the es to file block allocation trades data This populates the man aged account summary file 0371 w p0371 aggregate day trades to fi This wor
6. Business Field Column Name Date Type Definition Null Organization ORG_CD CHAR 20 Unique identifier for this organization No Organization Display Name ORG NM CHAR 60 Short name for the organization that is Yes used for display purposes Organization description ORG DESC TX CHAR 100 Description of this organization Yes Line Organization PRNT ORG CD CHAR 20 Identifies the parent organization of Yes which this organization is considered to be a child Modification Date MODFY DT DATE Identifies the last modified Date and Yes time Modified User MODFY ID NUMBER 10 Identifies the user id of the user who Yes last modified the data Comment COMMENT TX CHAR 4000 Comment Creating Organization in the Database through Excel Upload 2 Add records to the table by using a SOL script similar to the sample script in Figure 8 Sample SQL Script for Loading KDD ORG insert into KDD ORG ORG CD ORG NM ORG DESC TX PRNT ORG CD MODFY DT MODFY ID COMMENT TX values TestOrgA TestOrgA TestOrgA null null null null insert into KDD ORG ORG CD ORG NM ORG DESC TX PRNT ORG CD MODFY DT MODFY ID COMMENT TX values TestOrgB TestOrgB TestOrgB TestOrgA null null null insert into KDD ORG ORG CD ORG NM ORG DESC TX PRNT ORG CD MODFY DT MODFY ID COMMENT TX values TestOrgC TestOrgC TestOrgA null null null MODFY ID COMMENT TX TestOrgC MODFY
7. Fraud Detection Informatica Workflows The following sections describe the Informatica workflows that are required for deriving and aggregating data for Fraud Detection e Informatica Miscellaneous Optional Workflows Fraud Detection e Informatica Miscellaneous Pre Watch List Workflows Fraud Detection e Informatica Watch List Workflows Fraud Detection e Informatica Miscellaneous Post Watch List Workflows Fraud Detection e Informatica Summary Workflows Fraud Detection Informatica Miscellaneous Optional Workflows Fraud Detection Optional Miscellaneous workflows perform processing in support of workflows in multiple functional areas Figure 44 illustrates Optional Miscellaneous Informatica workflows for Fraud Detection Sequence 0 0070 2020 Note Processing can occur simultaneously within each sequence Workflows with predecessors must wait for predecessor results before workflow processing can begin Figure 44 Optional Workflows Fraud Detection The application completes these workflows in one sequence processing of the workflows does not require data from external predecessors 158 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Fraud Detection Informatica Workflows Chapter 5 Informatica Workflows Table 61 describes the Optional Miscellaneous workflows in Figure 44 Table 61 Optional Workflows Fraud Detection
8. sss 154 Informatica Summary Workflows Broker Compliance seen 155 Informatica Balances and Positions Workflows Broker Compliance ssssseeeenee 156 Informatica Miscellaneous Post Watch List Workflows Broker Compliance sse 157 Fraud Detection Informatica Workflows 5 eerte then h etie eben ege ded deseaba ton eee no 158 Informatica Miscellaneous Optional Workflows Fraud Detection sss 158 Informatica Miscellaneous Pre Watch List Workflows Fraud Detection sse 159 Informatica Watch List Workflows Praud Detection esses eene tette tenentes 160 Informatica Miscellaneous Post Watch List Workflows Fraud Detection sse 163 Informatica Summary Workflows Fraud Detection sesseeeeseee eene tentent 164 Insurance WOrk OWS osse cete ege ine pui ume etui eid m ee eum 166 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide ix Contents Informatica Miscellaneous Pre Watch List Workflows Insurance eee 166 Informatica Watch List Workflows Insurance eeeseeseeeeeee eene teneret ntn nte tenens 167 Informatica Miscellaneous Post Watch List Workflows Insurance eseeeeeeeeeneeetetnnen 169 Informatica Summary Workflows Insurance eee eene nennen nennen nene entere nennen 170 CHAPTER
9. Table 72 KDD AUTO CLOSE ALERT SCORE lt 75 and STATUS NW CLS_ OPRTN SET XPRSN ORDER ALERT AT NEXT OPRT ACTIVITY ID ID TR ID OPRTR CD VALUE TX N SET ID DMN CD CD CMMNT TX 100 2 113000022 75 101 MTS NULL NULL 101 NULL 113000008 NW NULL MTS NULL NULL Running the Auto Close Alert To Run Auto Close Alert Auto Close Alert is part of the Behavior Detection subsystem Oracle Financial Services provides default job templates and job template groups for running Auto Close Alert You can modify these jobs using the Administration Tools Refer to the Oracle Financial Services Behavior Detection Platform Administration Tools User Guide Release 6 1 3 for more information To run Auto Close Alert follow the steps 1 Verify that the dispatcher is running 2 Run the start_mantas sh script as follows start_mantas sh 506 where 506 is the job template that Oracle Financial Services provides to run the Auto Close algorithm 182 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Automatic Alert Suppression Chapter 6 Post Processing Tasks Automatic Alert Suppression The Alert Management subsystem provides actions that enable an analyst to specify that the system close a particular entity s alerts on a specific scenario automatically This is called Alert Suppression The system runs the Alert Suppression algorithm to close n
10. ERROR DESC TX Description of a resulting error or warning The system also uses the KDD CAL table to obtain information such as the dates of the last day of previous month and end of weeks Refer to Table 93 on page 239 for contents of the KDD CAL table Database Statistics Management For each of the MANTAS BUSINESS and MARKET schemas the system uses a script to manage Oracle database statistics These statistics determine the appropriate execution path for each database query Logs The 1og category RUN STORED PROCEDURE property controls logging for the process location entty in the INSTALL DIR database db tools mantas cfg categories cfg file Using Database Statistics Management The system calls each script as part of nightly processing at the appropriate time and with the appropriate parameters e MANTAS Schema analyze mantas sh TABLE NAME analysis type e BUSINESS Schema analyze business sh TABLE NAME analysis type e MARKET Schema analyze market sh TABLE NAME analysis type The analysis type parameter can have one of the following values e DLY POST LOAD Use this value to update statistics on tables that the system just loaded for BUSINESS and MARKET schemas e ALL Use this once per week on all schemas Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 251 Database Statistics Management Chapter 7 Batch Processing Utilities
11. Sunday 1 Monday 2 Tuesday 3 Saturday 7 If the table does not contain records for any calendar in the list the system logs a Warning message that the specific calendar contains no weekly off days 214 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Alert Purge Utility Chapter 7 Batch Processing Utilities Alert Purge Utility Directory Structure Occasionally ingestion of certain data results in the creation of false matches alerts and activities While correction and data re ingestion is possible the system does not remove these erroneously generated matches alerts and activities automatically The Alert Purge Utility enables you to identify and remove such matches alerts and activities selectively based on the Behavior Detection Job ID or Behavior Detection Scenario ID and a date range with optional alert status codes Additional parameters enable you to simulate a purge run to determine all found matches alerts and activities using the input parameters You can also limit the alerts in the purge process only to those that contain false matches The utility consists of a UNIX shell script Java executables and a configuration file in which you define the process parameters to use in the purge processing The system directs output to a configurable log file processing appends this log with information about subsequent executions of the scripts This section cov
12. Sequence Workflow Number Folder Number Workflow Name Predecessor 3 ORION Common 3000 w ph3000 Adjust WL WLS ORION Common 2180 3 ORION Common 3005 w ph3005 apply cust KYC risk 3 ORION Common 3030 w ph3030 truncate wls 4 ORION Common 3010 w ph3010 truncate Is 4 ORION Common 3020 w ph3020 truncate nms 4 ORION Common 3040 w ph3040 truncate wls2 4 MLM Brokerage C 3041 w ph3041 create addresses from inst ommon ORION Common 3051 w ph3051 create addresses from fotps 4 ORION Common 3090 w ph3090 create external entities from fotps ORION Common 2190 ORION Common 3091 4 MLM Brokerage C 3120 w ph3120 create external entities from inst MLM Brokerage Common 2192 ommon 5 ORION Common 3070 w ph3070 load watch list staging table ORION Common 3000 ORION_Common 3030 5 ORION_Common_ 3140 w ph3140 write fotps associations to Is ORION Common 3010 ORION Common 3051 ORION Common 3090 5 MLM Brokerage C 3160 w ph3160 write inst associations to Is ORION Common 3010 ommon MLM Brokerage Com mon 3041 MLM Brokerage Common 3120 6 ORION Common 3100 w ph3100 load staging fuzzy matches ORION Common 3020 ORION Common 3070 ORION Common 3090 6 ORION Common 3180 w ph3180 write Is to link tables ORION Common 3110 ORION Common 3140 MLM Brokerage Common 3160 7 ORION Common 3150 w ph3150 load staging and validate watch list ORION Common 3040 ORION Common 3070 ORION Common 3090 ORION Common 3 091 ORION Common 3100
13. Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 15 Performing Dispatcher Tasks Chapter 2 Behavior Detection Jobs Table 7 shows operating system variables in the system env file Table 7 Operating System Environment Variables in system env File Variable Description PATH Augmented to include KDD_HOME bin and the SORACLE HOME SORACLE HOME bin pair for Oracle LD LIBRARY PATH LIBPATH Augmented to include KDD HOME 1ib and SHLIB PATH based on operat SORACLE HOME 1ib for Oracle ing system Starting the Dispatcher Although multiple jobs and mantas instances can run concurrently in the application only one dispatcher service per database per installation should run at one time The application provides a script to check on the status of the dispatcher automatically and restart it if necessary Oracle Financial Services recommends this method of running the dispatcher To start the dispatcher follow the steps 1 Verify that the dispatcher is not already running by typing ps ef grep dispatch and pressing Enter at the system prompt If the dispatcher is running an instance of the dispatcher appears on the screen for the server If the dispatcher is not running proceed to Step 2 2 Type start_chkdisp sh lt sleep time gt and press Enter at the system prompt to start the dispatcher The dispatcher queries the database to check for any new jo
14. DB TOOLS refresh temp table sh TMP HIDREL NT WIRE VW DB TOOLS refresh temp table sh TMP HIDREL NT ACTAXID VW DB TOOLS refresh temp table sh TMP HIDREL NT ACADDR VW DB TOOLS refresh temp table sh TMP HIDREL NT ACPHONE VW DB TOOLS refresh temp table sh TMP HIDREL NT ACEMAIL VW DB TOOLS refresh temp table sh TMP HIDREL NT ACPSWRD VW DB TOOLS refresh temp table sh TMP HIDREL NT INST VW SDB TOOLS refresh temp table sh TMP HIDREL NT WIREACBENE VW DB TOOLS refresh temp table sh TMP HIDREL NT WIREACORIG VW SDB TOOLS refresh temp table sh TMP HIDREL NT CUACTAXID VW DB TOOLS refresh temp table sh TMP HIDREL NT CUACADDR VW DB TOOLS refresh temp table sh TMP HIDREL NT CUACPHONE VW SDB TOOLS refresh temp table sh T P HIDRE P HIDRE P HIDRE P HIDRE P HI DRE P HI DRE TMP HIDREL NT WIRE TMP HIDREL NT ACTAXID P HIDREL NT ACADDR L NT ACPHONE L NT ACEMAIL L NT ACPSWRD HIDREL NT INST HIDREL NT WIREACBENE HIDREL NT WIREACORIG L NT CUACTAXID L NT CUACADDR L NT CUACPHONE TMP HIDREL NT CUACEMAIL Execute the link analysis network generation job The product job template ID 3 Execute the scenario job The product job template ID is 116200024 To populate the temporary tables for ML NetworkOfAcEn fAC scenario follow the steps 1 Execute these refresh temporary table processes these commands can be run in parallel DB TOOLS re DB TOOLS re 9g
15. Trace SQL exception Set to true for SQL tracing verbose to trace low level JDBC calls com sra kdd tools database debug true Specify which priorities are enabled in a hierarchical fashion i e if DIAGNOSTIC priority is enabled NOTICE WARN and FATAL are also enabled but TRACE is not Uncomment the desired log level to turn on appropriate level s Note DIAGNOSTIC logging is used to log database statements and will slow down performance Only turn on if you need to see the SQL statements being executed TRACE logging is used for debugging during development Also only turn on TRACE if needed HHH HH HH H fatal true warning true notice true diagnostic false trace false time zone US Eastern Specify whether logging for a particular level should be performed synchronously or asynchronously Continued on next page Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 205 About Administrative Utilities Chapter 7 Batch Processing Utilities Continued from previous page log fatal synchronous false log warning synchronous false log notice synchronous false diagnostic synchronous false trace synchronous true Specify the format of the log output Can be modified according to the format Specifications at http logging apache org log4j docs api org apache 10og4j PatternLayout html NOTE Because of the nature of asynchronous logging detailed i
16. categories cfg File In the lt INSTALL_DIR gt database db_tools mantas_cfg categories cfg file you can modify the default location to where you want to direct logging output for each utility The entries that you make require a specific format the file contains instructions and examples of correct formatting Figure 55 provides a sample categories cfg file db db Gb HH Gb Gb dt db db db db HH HK dt db HH Gb Gb HF FH Common Logging categories configuration for Oracle Financial Services Database Specify the log location for messages of a specific category The property name should be of the form log category CATEGORY NAME location If logging to a category that is not specified below the messages are logged to a configurable default location Valid values are console syslog eventviewer mantaslog an e mail address or the full path to a file If specifying mantaslog also specify the property log mantaslog location with the desired path to the logfile in install cfg If running the algorithms use the format job job gt datetimestamp for the mantaslog filename For other subsystems the format is mantaslog datetimestamp NOTE Category names cannot contain the following reserved words fatal warning notice diagnostic trace category or location List multiple locations for each property by using a comma delimiter NOTE These are commented out because Oracle Financial Services does not currently route
17. eerte trinitate tero Reed 233 Table 91 Case Metadata s erp b e UE oe RARE RE EEE 235 Table 92 Calendar Manager Utility Directory Structure 236 Table 93 KDD CAL Table Contents esee nne ene 239 Table 94 Data Retention Manager Directory Structure sss 241 Table 95 Data Retention Manager Processing Parametets sss 245 Table 96 Partition Name Formats rettet locii i 247 Table 97 BUSINESS KDD DR MAINT OPRTN Table Contents 249 Table 98 BUSINESS KDD DR JOB Table Contents see 250 Table 99 BUSINESS KDD DR RUN Table Contents 251 Table 100 Configuration Instructions for the install cfgfFile 263 Table 101 XML Code Operators sesiis siiis iasi 264 Table 102 Data Analysis Tool XML Input Files een 2n Table 103 Command Line Arguments sse etn 272 Table 104 Data Analysis Report Output eene 273 Table 105 Troubleshooting Data Analysis Tool Errors sese 273 Table 106 Get Dataset Query Variables seen 274 Table 107 General Scenario Migration Parameters sse 278 Table 108 Scenario Extraction Parameters sss 279 Table 109 Scenario Load Parametets ettet tiet triti 280 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide xix List of Tables Table 110 Table 111 Table 112 Table 113 Table
18. Alert Purge Utility Chapter 7 Batch Processing Utilities Table 80 Alert Purge Utility Parameters Continued Parameter Description start_date Indicates the start date for the Scenario ID when the scenario parameter is in use in the format DD MON YYYY HH MM SS or DD MON YYYY When using only the date the time component defaults to midnight You must set this parameter to NULL if it is not used However when using the scenario parameter it cannot be set to NULL end_date Indicates the end date for the Scenario ID when the scenario parameter is in use in the for mat DD MON YYYY HH MM SS or DD MON YYYY When using only the date the time component defaults to midnight You must set this parameter to NULL if it is not used However when using the scenario parameter it cannot be set to NULL alert_status Identifies an alert status code when the scenario parameter is in use against which to restrict the Alert Purge Utility further Comma separated list Alert status codes include NW New OP Open CL Closed FL RO and RA When using the scenario parameter the alert_status must be used however you can set it to NULL Executing the Alert Purge Utility Processing for Purging To execute the Alert Purge Utility follow the steps 1 Verify that the Behavior Detection database is operational tnsping lt database instance name gt 2 Verify that th
19. Data Ingest Properties Table 31 Datalngest properties File Configuration Parameters Table 31 describes the parameters for the DataIngest properties configuration Configuration File file Property Name Description Example DB Connection URL Database URL for JDBC connections made by Java ingestion components The content and format of this value is specific to the database vendor and the vendor database driver jdbc oracle oci8 D109L2 DB Connection Server Database server on which the database software is executing This parameter is required in some circumstances where the database URL is not sufficient for the database driver software to connect to the database db1 clientname com DB Connection Instance Database instance to connect to on the database servers Typically the instance name matches the database name portion of the DB Connection URL D109L2 DB Connection User Database user name that Java ingestion components uses when connecting to the database The database user must have been assigned the appropriate privileges that Data Ingestion requires for interacting with the database INGEST_USER DB Connection Password Password that Java Ingestion components use when connecting with the database This is set by the Password Manager Utility DB Type The type of database being used Oracle MANTAS DBSchema Schema name for the Mantas database schema Ty
20. Running Trading Compliance Solution Data Loaders When FDT and MDT processing complete the system executes the runDL sh script for TCS trading and market data load files This activity loads data from the preprocessors and transformers into the FSDM schema The FullRefresh parameter in DataIngest xm1 controls use of full refresh or delta mode A value of true implies use of Full Refresh mode a value of alse implies use of Delta mode Setting of the default can be to one or the other overriding the default is possible for individual file types For reference data that is any file that has a load operation of Overwrite which the DIS specifies two options are available for loading data e Full Refresh Truncates the entire table before loading the file Use this mode when you plan to provide a complete set of records daily You must set the FullRefresh parameter in DataIngest xml to lt true gt to use the Full Refresh mode e Delta Mode Updates existing data and inserts new data Use this mode only when you plan to provide new or changed records daily You must set the FullRefresh parameter in DataIngest xml to lt false gt to use the Delta mode Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 67 Process Flow Chapter 4 Data Ingestion The system executes data loaders using the runDL sh script the following provides a sample command lt INSTALL_DIR gt ingestion_man
21. 1 Access the directory where the shell script resides cd INSTALL DIR gt database db tools bin 2 Start the batch shell script with the parameters in Table 95 run drm utility sh drm operation drm partition type drm owner drm object name drm weekly proc fl Script Examples The following are examples of running the script e Torun the utility for all daily tables in the BUSINESS schema execute the script run drm utility sh P D BUSINESS A N e To run the utility to drop a monthly partition of the BUSINESS table ACCT SMRY MNTH execute the script as follows using the same parameters as in the previous example run drm utility sh DM M BUSINESS ACCT SMRY MNTH N 246 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Retention Manager Chapter 7 Batch Processing Utilities Creating Partitions Maintaining Partitions When creating partition names use the formats in Table 96 Table 96 Partition Name Formats Partition Type Format and Description Monthly PYYYYMM where YYYY is the four digit year and MM is the two digit month for the data in the partition For example Data for November 2006 resides in partition P200611 Note The Data Retention Manager uses information in the KDD_CAL table to determine end of week and end of month boundary dates Weekly or Daily PYYYYMMDD where YYYY is the four digit year MM is the two digit month
22. DLY POST HDC Use this value to update statistics of the alert related archived data in _ARC tables in the BUSINESS and MARKET schema tables that the Behavior Detection UI uses to display alerts DLY_PRE_HDC Use this value to update statistics of the Mantas schema tables that contain the alert related information Note It is recommended that you do not modify the tables for DLY POST HDC and DLY PRE HDC The Behavior Detection Historical Data Copy procedures use these tables to archive alert related data DLY_POST_LINK Use this value to update statistics of the Mantas schema tables that contain network analysis information Run this option at the conclusion of the network analysis batch process The TABLE_NAME patameter optionally enables you to analyze one table at a time This allows scheduling of the batch at a more granular level analyzing each table as processing completes instead of waiting for all tables to complete before running the analysis process The metadata in the KDD_ANALYZE_PARAM table drive these processes For each table in the three schemas this table provides information about the method of updating the statistics that you should use for each analysis type Valid methods include EST STATS Performs a standard statistics estimate on the table EST PART STATS Estimates statistics on only the newest partition in the table Note For the EST STATS and EST PART STATS parameters the default s
23. Security Specifies properties used to produce security reference data Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 91 Data Ingestion Directory Structure Chapter 4 Data Ingestion Table 32 Datalngest xml File Configuration Parameters Continued Property Name Description Example Security AdditionalColumns Specifies additional columns of data that SCRTY SHRT NM ingestion components need to populate SCRTY ISIN ID when manufacturing security records PROD CTGRY CD PROD TYPE CD PROD SUB TYPE CD Symbol Specifies properties used for looking up security reference data by security short name Symbol DbTableName Specifies the name of the database table to SCRTY use when looking up security records by security short name Symbol KeyColumn Specifies the column name to use when SCRTY SHRT NM looking up security records by security short name Symbol ValueColumn Specifies the column to use for retrieving the SCRTY INTRL ID Behavior Detection assigned identifier for a security Symbol Category Specifies the category of data for the security BUSINESS table The category is a key for mapping to the database schema in which the security table resides SecurityISIN Specifies properties used for looking up security ISINs SecurityISIN DbTableName Specifies the name of the table to use when SCRTY looking up a security using
24. The preprocessor moves files in the inbox directory All other components move their input files to date labeled subdirectories within the backup directory Periodically an Oracle Financial Services client can run the runIMC sh script to perform the Ingestion Manager cleanup activities This script deletes old files from the archive area based on a configurable retention date Periodic running of the cleanup script ensures that archive space is available to archive more recent data 112 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Ingestion Archiving Chapter 4 Data Ingestion Archiving Database Information The database archiving process is explained in Figure 25 1 Data ingestion inserts data into the newest partition Daily Partitions 2 System generates alerts for detected behavior zzgl 3 System copies alert data to ARC tables Weekly Partitions Historical Data Copy User Interface Alert information from archive tables Tab pages display current information Figure 25 Database Archiving Process The Data Ingestion subsystem uses the following procedure 1 Processing places data in the newest partition of the partitioned tables 2 Scenarios examine the data in the partitioned tables the system then generates alerts for detected behaviors 3 Historical Data Copy processing copies the information that generated alerts reference to th
25. MDS receives incoming market data and makes it available for transformers R we cy o Reference Data Transaction Data Reference Data Market Data Enriched Transaction Data FDT enriches transaction Reference Data data Transaction Transaction Data Reference Data Preprocessors error check input files assign sequence numbers and resolve cross references Legend E Java C Informatica Figure 11 Data Ingestion Subsystem MDT enriches market data Load and Transfor Data Loaders batch insert and update reference transaction and market data Enriched Market Data gt a x hem k Y dil M Sc E d ta FSD j Java and security provide portability in performing database support functions index rebuilds and analysis summary calculations Informatica is used to enrich data after it is loaded into the database uonsoeBu ejeq r 1o3deu5 MO SS920Jd 54 Process Flow Chapter 4 Data Ingestion Data Ingestion Process Summary Figure 12 provides a high level view of the Data Ingestion process for Oracle Financial Services Trading Compliance Solution TC Anti Money Laundering AML Broker Compliance Solution BC Fraud FR and Insurance Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Process Flow Chapter 4 Data Ingestion f Refer to Table 17 for data types by group These steps only need to
26. 19 Starting Jobs without the Dispatchet eceo neret rrt ee rires oer R pe eee EA 19 Restat no a Moe EE 20 Restarting Jobs without the Dispatches sess inneni naa aa aias 21 SCOPPINS NODS edd S 21 Monitoring and Diagnosing Jobs ttt ettet rrt titi csssotsosudsebosveseenssstioneeusboeveseessvs es 22 Clearing Out the System Logs miti ettet o Hn Ee ER EXE E e HEB E EP XEEREXER ERE E EBEN MEER EEE 23 Clearing the Dispatch L9 nemore ete ert edu ine eds Qu EO e dr Tans Rd ue A ER 23 Cleaning the Job EOgs s iore retento rere e FREE RR EHE RO EUER OE A R REANA REER 23 Recovering Jobs from a System Gras Kisera Io oosis er dE E PIRE ETUR aaa eot end Re echo nouo Rot eoa 24 CHAPTER 3 Security Configuration ssssssssseeeee 25 About the Oracle Financial Services User Authentication sse eene 25 Undetstaniding SMS M eissi nseni ES E EENS ESKE aars i a ni A 26 Accessing Oracle Financial Services Behavior Detection 26 S tttng p a SEE testen teer tatit bee echa Ete be E 27 User Group and User Roles eri bee bend aede tei tiere estie des deett eiie etie ana 27 Mapping a User toza Single Usep Group sis s e rettet ep d nas R EE GEE TERVEET 30 Mapping a user to Multiple User Groups within Alert Management and Case Management eese 30 Mapping a user to Multiple User Groups across Alert Management Case Management and Other Applications 30 Mapping a Function toa Role ceca tnn ene teet
27. INSERT INTO KDD CAL HOLIDAY CLNDR NM CLNDR DT HLDY NM HLDY TYPE CD VALUES SYSCAL TO DATE 09 04 2006 MM DD YYYY Labor Day 2006 C INSERT INTO KDD CAL HOLIDAY CLNDR NM CLNDR DT HLDY NM HLDY TYPE CD VALUES SYSCAL TO DATE 11 22 2006 MM DD YYYY Thanksgiving Day 2006 C INSERT INTO KDD CAL HOLIDAY CLNDR NM CLNDR DT HLDY NM HLDY TYPE CD VALUES SYSCAL TO DATE 12 25 2006 MM DD YYYY Christmas Day 2006 C COMMIT Figure 57 Sample KDD CAL HOLIDAY Table Loading Script 212 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide About Annual Activities Chapter 7 Batch Processing Utilities Table 77 provides the contents of the KbD CAL HOLIDAY table Table 77 KDD CAL HOLIDAY Table Contents Column Name Description CLNDR NM Specific calendar name CLNDR DT Date that is a holiday HLDY NM Holiday name for example Thanksgiving or Christmas HLDY TYPE CD Indicates whether the business is Closed C or Shortened S SESSN OPN TM Indicates the opening time of the trading session for a short ened day The format is HHMM SESSN CLS TM Indicates the closing time of the trading session for a shortened day The format is HHMM SESSN TM OFFSET Indicates the timezone offset for SESSN OPN TM and TX SESSN CLS TM When the system runs the set mantas date sh script it queries th
28. Loading Holidays Typically on an annual basis you populate holidays for the upcoming calendar year into the Behavior Detection KDD CAL HOLIDAY database table This ensures that the table contains holidays for at least the next year Figure 57 provides an example of a SQL script for loading the table Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 211 About Annual Activities Chapter 7 Batch Processing Utilities INSERT INTO KDD CAL HOLIDAY CLNDR NM CLNDR DT HLDY NM HLDY TYPE CD VALUES SYSCAL TO DATE 01 01 2006 MM DD YYYY New Year s Day 2006 C INSERT INTO KDD CAL HOLIDAY CLNDR NM CLNDR DT HLDY NM HLDY TYPE CD VALUES SYSCAL TO DATE 01 16 2006 MM DD YYYY Martin Luther King Jr s Birthday 2006 C INSERT INTO KDD CAL HOLIDAY CLNDR NM CLNDR DT HLDY NM HLDY TYPE CD VALUES SYSCAL TO DATE 02 20 2006 MM DD YYYY President s Day 2006 C INSERT INTO KDD CAL HOLIDAY CLNDR NM CLNDR DT HLDY NM HLDY TYPE CD VALUES SYSCAL TO DATE 04 14 2006 MM DD YYYY Good Friday 2006 C INSERT INTO KDD CAL HOLIDAY CLNDR NM CLNDR DT HLDY NM HLDY TYPE CD VALUES SYSCAL TO DATE 05 29 2006 MM DD YYYY Memorial Day 2006 C INSERT INTO KDD CAL HOLIDAY CLNDR NM CLNDR DT HLDY NM HLDY TYPE CD VALUES SYSCAL TO DATE 07 04 2006 MM DD YYYY Independence Day 2006 C
29. Miscellaneous Utilities Chapter 4 Data Ingestion this Account Address record Futhermore in the case of Account Address it is not just the Account Identifier of an Account Address record that is of interest The Address Purpose is also of interest So when we look in the database for Account Address records that match a given Account Address record in a DIS file we look to match both the Account Identifier field and the Address Purpose field This processing is controlled by configuration parameters in the ChangeLog section of DataIngest xml There is a Default subsection that defines parameters that are relevant to all data types Each data type have their own subsection that defines relevant parameters as well as the fields that are checked for changes Table 40 lists the Change Log parameters Table 40 Change Log Parameters Parameter Description class The Java class name to perform change log processing For the first type of processing defined above ChangeLog should be used For the second type SetChangeLog should be used seqBatchSize Specifies the batch size when retrieving sequence ids for creating ChangeLog records enabled Specifies whether ChangeLog processing should be done for the given data type queryKey This is only relevant for the second type of processing that is when class SetChangeLog This defines the key that is used to query for reference data records matching the given one In the
30. Table 72 Table 73 Table 74 Table 75 Table 76 Table 77 Table 78 Table 79 Table 80 Table 81 Table 82 Table 83 Table 84 Table 85 Table 86 Commonly Used Alert Closing Attributes sse 179 KDD AUTO CLOSE ALERT AGE gt 30 sse 182 KDD AUTO CLOSE ALERT SCORE lt 75 and STATUS NW 182 HDC Configurable P rametets ettet tette ttttntrtn toin 186 KDD BUS NTITY PATH Metadata Table sees 187 KDD BUS NTITY PATH CFG Metadata Table 188 AAZPTASCHpIBus edd p NM I a MR Sod AE 193 KDD CAL HOLIDAY Table Contents eeeeeeeeenren 213 KDO CAL WKLY OFF Table Contents sss 214 Alert Purge Utility Directory Structure sese 215 Alert Purge Utility PatametetS i anneren isi i a E ER 218 Example of Matches and Alerts Associated with Purge Alerts 221 Batch Control Utility Directory Structute sse 224 KDD PRCSNG BATCH Table Contents see 226 Sample KDD PRCSNG BATCH Table with Single Batch 226 Sample KDD PRCSNG BATCH Table with Intra day Processing 226 Sample KDD PRCSNG BATCH Table with Multiple Country Processing 227 Table 87 KDD PRCSNG BATCH CONTROL Table Contents 228 Table 88 KDD PRCSNG BATCH HIST Table Contents eee 228 Table 89 Directory Components tentent 231 Table 90 Alett Metadata
31. e Request Assignment Determines the group of users responsible for handling account approval requests and pre trade approval requests 192 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Account Approval Pre Trade Approval Tasks Chapter 6 Post Processing Tasks For this purpose the following scripts which is part of the Database Tools that resides in the lt INSTALL_DIR gt database db_tools bin directory call specific database procedures Table 76 AA PTA Scripts Task Load Employee User Groups Script run apprvl load emp ug Database Procedure Description P PTA EMP UG POPULATION sh Maps employees within the Oracle client to the AA PTA pre packaged Employee User Group CREMPLOYEEUG All employee users provided in the System Logon file with the Source System set to MTS are mapped to CREMPLOYEEUG which gets captured in the CSSMS_USR_GROUP_MAP table This gives an employee access to the AA PTA application To identify all the employees that must have access to AA PTA the Employee and System Logon files must be provided The system logon IDs for all employees within the Oracle client captured in the System Logon file must have the Source System SRC_SYS_CD field value set to MTS Refer to the Oracle Financial Service Behavior Detection Platform Data Interface Specification for details on how to populate the Employee and System Logon files Request Assi
32. limit matches Y purgezN batch size 5000 job null Scenario null Enter dates with quotes in the following format DD MMM YYYY HH MI SS or DD MON YYYY start date null end date null alert status NW Base Working Directory required to put the temporary log from the Database server ap storedproc logdir tmp Figure 59 Configuration Information Note Not specifying a value of nul for example leaving a value blank in this section of the install cfg file causes undesirable results Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 217 Alert Purge Utility Chapter 7 Batch Processing Utilities Table 80 describes required and optional parameters for this utility Table 80 Alert Purge Utility Parameters Parameter Description purge Determines how the utility performs processing depending on the specified value N default Performs all processing up to the point of the purge The utility identifies resulting matches alerts and actions but performs no purging Y Performs the above in addition to purging matches alerts and actions limit matches Identifies restrictions on the matches to delete e Y default If a match that you want to delete is part of an alert that contains matches that you do not want to delete do not delete this match either applies to multi match alerts N Deletes all selected matches for purgi
33. Batch Start Status 229 KDD PRSCNG BATCH HIST Table Batch End Status 229 Database Partitioning Process wesc esseeesseesesesnssestssestssestsseseaey 243 Sample install cfg File for Scenario Migration sss 277 Sample install cfg File for Alert Correlation Rule Migration 289 Sample Logging Configuration File sse 306 xvi Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide List of Tables Table 1 Conventions Used in this Guide sss XXV Table 2 Subsystems and their Components sss 4 Table 3 Shell Scripts that Call mantas Processes sse 12 Table 4 KDD JOB TEMPLATE with Sample Job Template Group 12 Table 5 Environment Variables in system env File eee 15 Table 6 Database Environment Variables in system env File eee sss 15 Table 7 Operating System Environment Variables in system env File 16 Table 8 Alert Management Roles and User Groups sse 28 Table 9 Case Management Roles and User Gtoups sse 28 Table 10 Account Approval Pre Trade Approval Roles and User Groups 29 Table 11 Relationships between Data Points sss 32 Table 12 Access Control Items and Locations sse 33 Table 13 KDD_JRSDCN Table Attrib tes ss aaa arnitiai ires ttertacilati t
34. DateActionTaken KDD ACTIVITY START DT UserNameTakingAction KDD REVIEW OWNER OWNER ID OwnerUserName KDD REVIEW OWNER OWNER ID AlertStatus KDD REVIEW STATUS STATUS CD Element Scenarios ScenarioName KDD SCNRO SCNRO SHORT NM ScenarioCataloglD KDD SCNRO SCNRO CTLG ID Covers all actions taken on alert not just the closing action 234 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Batch Export Utility Chapter 7 Batch Processing Utilities Note Case metadata xsd lt alert mgmt gt WEB INF classes conf ui_config dtd mantasCaseExportDa ta xsd Table 91 Case Metadata Case Metadata System Name PDF Hard coded to Oracle Financial Services Behavior Detection File name Element ExportDate System Date Element CaselD KDD REVIEW REVIEW ID Element CreatorUserName KDD REVIEW OWNER OWNER ID Element CreationDate KDD REVIEW CREAT TS Element CurrentStatus KDD REVIEW STATUS STATUS CD Element FocalEntityDisplayID KDD REVIEW FOCAL NTITY DSPLY ID Element FocalEntityName KDD REVIEW FOCAL NTITY DSPLY NM Element FocalTypeCode KDD CENTRICITY CNTRY TYPE CD Element CaseType KDD CASE TYPE CASE TYPE NM Element CaseSubType KDD CASE TYPE CASE SUB TYPE NM El
35. Detection uses inbuilt SMS for its authentication and authorization The SMS has a set of database tables which store information about user authentication Installation of 128 bit encryption support from Microsoft can secure the Web browser Oracle Financial Services encourages using the Secure Socket Layer SSL between the Web browser and Web server for login transaction While the Web Application server uses a browser cookie to track a user s session this cookie is temporary and resides only in browser memory When the user closes the browser the system deletes the cookie automatically The application uses Advanced Encryption Standard AES security to encrypt passwords that reside in database tables in the configuration schema on the database server and also encrypts the passwords that reside in configuration files on the server Analysts HTTP or HTTPS Web Browser Via Weblogic s NS API plug in 1 Web Application Sener HTTP or HTTPS Communication SMS Via NSAPI plug in Party EAM Tool Extemal uth Repository Figure 3 Oracle Financial Services Architecture Security View 6 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide About Oracle Financial Services Operations Chapter 1 The Oracle Financial Services Behavior Detection Platform The EAM tool is an optional third party pluggable component of the security view The tool
36. MLM Brokerage Common 3120 MLM Banking Common 3130 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 139 AML Brokerage Workflows Chapter 5 Informatica Workflows 8 ORION Common 3170 w ph3170 update staging list risk ORION Common 3150 9 ORION Common 3190 w ph3190 apply risk to nonacct entities ORION Common 3005 ORION_Common 3170 9 ORION Common 3200 w ph3200 apply membership to entities ORION Common 3170 10 MLM Brokerage C 3171 w ph3171 update account customer risk ORION Common 3190 ommon ORION_Common 3200 11 ORION Common 3191 w ph3191 apply risk to acct entities MLM Brokerage Common 3171 ORION Comm 0on 3190 12 ORION Common 3210 w ph3210 update fotps activity risk ORION Common 3191 12 MLM Brokerage C 3220 w ph3220 update bot activity risk ORION Common 3191 ommon 12 MLM Brokerage C 3230 w ph3230 update inst activity risk ORION Common 3191 ommon Note If you are running any of these combination you need to run workflows 3005 as well as 3171 e Oracle Finanical Services Behavior Detection AML and KYC e Oracle Finanical Services Behavior Detection Fraud and KYC Oracle Finanical Services Behavior Detection AML Fraud and KYC Informatica Miscellaneous Post Watch List Workflows AML Brokerage Post Watch List Miscellaneous workflows perform processing in support of workflows in multipl
37. MatchedEntity SecurityMarketDaily Security AccountToPeerGroup MarketIndexMemberSecurity AccountSupplementalAttribute CustomerSupplementalAttribute WatchListEntry RestrictionList AutomatedQuote LoanProduct MailHandlingInstructionActivity OrganizationToMortgageType ReferenceTableDetail ServiceVendor EnergyAndCommodityTrade LoanOriginationProduct Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Process Flow Chapter 4 Data Ingestion Table 19 Data Files by Group Continued Group Data Files 4 AccountAddress CustomerToProductsOffered AccountAssetAllocation CustomerToMarketsServed AccountBalance EmployeeAddress AccountCollateral EmployeeEmailAddress AccountFeature EmployeePhone AccountGroupMember EmployeeToAccount AccountlnvestmentObjective EmployeeToOrganization AccountPosition EmployeeTradingRestriction AccountPositionPair FirmAccountPosition AccountToCorrespondent Fraud AccountToCustomer MarketNewsEvent AccountToOrganization MutualFundBreakpoint AnticipatoryProfile SecurityGroupMember AccountProfitAndLoss SecurityInvestmentRating ControllingCustomer SecuritySelectListEntry CustomerAddress SecurityTradingRestriction CustomerBalance StructuredDeal CustomerCountry SystemLogon CustomerEmailAddress OnlineAccountToAccount CustomerPhone ManagedAccount CustomerToCustomerRelationship AccountProfileStage AccountScheduledEvent EnergyAndCommodityFirmDaily LoanOrigination Ener
38. NULL the system uses the default auto close activity type code 178 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Auto Close Chapter 6 Post Processing Tasks To Set Up Auto Close Rules e The CMMNT_TX column specifies an optional text comment to associate with an alert that is closed by this rule The Auto Close Alert algorithm does not close a locked alert The system locks an alert when an analyst investigates it and then unlocks it when the analyst releases it All locked alerts are skipped until the next time the Auto Close Alert algorithm is run The Oracle Financial Services administrator must fill in rows in the KDD_AUTO_CLOSE_ALERT table with the criteria for auto closing the alerts The system uses the KDD_REVIEW table to provide available attributes for use in the Auto Close algorithm Refer to the Oracle Financial Services Behavior Detection Platform FSDM Reference Guide Volume 2 Release 6 1 3 for more information about the KDD_REVIEW table To set up auto close rules follow the steps 1 Formulate the criteria for auto closing alerts using the attributes in the Alert Closing Attributes KDD_AUTO_CLOSE_ALERT table The Alert Identifier ALERT_ATTR_ID column is needed later in this set of instructions Table 70 describes commonly used Alert Closing Attributes Table 70 Commonly Used Alert Closing Attributes Alert Attribute Alert Ide
39. Note You should ensure that all schema names that is MANTAS BUSINESS and MARKET ate in uppercase 198 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide About Administrative Utilities Chapter 7 Batch Processing Utilities This configuration file supports the following database utilities Calendar Mangager Batch Control Truncate Manager Scenario Migration Alert Purge Data Retention Manager Email Notification Data Analysis Tool H H HE HEHE Gb HHHH file contains some properties that are common and specific properties for each the tools H OH rho oO HHHHHHHHHHHHHHEH COMMON CONFIGURATION ENTRIES H H H HHHHHHHHHHHHHHHHH database driverName oracle jdbc driver OracleDriver utils database urlName jdbc oracle oci Ti5010S10 utils database username DB UTIL USER TEST58 utils database password DB UTIL USER TEST58 Schema mantas owner mantas TEST58 utils miner user KDD MNR TEST58 utils miner password utils altio username KDD ALTIO TEST58 schema business owner BUSINESS TEST58 schema market owner MARKET TEST58 utils data directory users mantast Solaris10 mantas58 b09 Ti5010S10 Iron 13080 WAS databa se db tools data ingest user INGEST USER TEST58 ingest password THHHHHHHHHHHHHEHEE CALENDAR MANAGER CONFIGURATION H H HHHHHHHHHEH The look back and look forward days of the provided date These values are required to update the KDD CAL table The m
40. Table 32 Datalngest xml File Configuration Parameters Continued Property Name Description Example SSAParams Population SSAParams PersonalKeyLevel Specifies the Population rule set to be used This generally corresponds to a Country Language rule set for example Australia Brazil UK and USA The name of the population corresponds to the name of the file provided by Informatica Identity Resolution that contains the rules for matching names in the given language Specifies the Key Level to be used when generating Informatica Identity Resolution Keys for personal names Standard Keys overcome more variation than Limited Keys while using less disk space than Extended Keys aml ysp Standard Extended or Limited SSAParams BusinessKeyLevel Specifies the Key Level to be used when generating Informatica Identity Resolution Keys for business names Standard Keys overcome more variation than Limited Keys while using less disk space than Extended Keys Standard Extended or Limited SSAParams 1 PersonalSearchLeve Used in defining the type of Search Strategy to use when searching for personal names The four possible values allow adjustment to the thoroughness of the search The wider the search the more candidates are typically returned which may increase the reliability of the search however it uses more resources and take longer Narrow Exhaustive Typical or Extreme SSAPara
41. This workflow maintains the External Entity table It from fotps derives the entities from the Front Office transaction table 322 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Informatica Workflows Appendix D Informatica Workflow Details Table 120 Informatica Workflow Description Continued Workflow Number Workflow Name Description 3091 w ph3091 create PartyNmEE from fot This workflow maintains the names in External Entities ps from FOTPS 3100 w ph3100 load staging fuzzy matche This workflow is a wrapper for the fuzzy matching map S pings and scripts 3110 w ph3110 write INS associations to This workflow loads the Link Stage with any entity asso ls ciations from INSURANCE 3140 w ph3140 write fotps associations This workflow loads the Link Stage with any entity asso to 1s ciations from FOTPS 3150 w ph3150 load staging and validate This workflow validates entities and their membership watch list lists It starts with WLS and creates records in WL S2 3170 w ph3170 update staging list risk This workflow defines the highest risk for each entity based on various sources 3180 w ph3180 write 1s to link tables This workflow writes link stage associations to various link tables 3190 w ph3190 apply risk to nonacct ent This workflow applies all risk related values to all entity ities tab
42. and DD is either the date of the data daily or the date of the following Friday weekly for the data in the partition For example Data for November 30 2006 resides in partition P20061130 Data for the week of November 19 November 23 2006 resides in partition P20061123 Note The Data Retention Manager uses information in the KDD_CAL table to determine end of week and end of month boundary dates Note Data Retention Manager assesses the current status of partitions on the specified table to determine the requested partition If the system previously fulfilled the request it logs a warning message Data Retention Manager does not support multiple partition types on a single table If an Oracle Financial Services client wants to alter the partitioning scheme on a table that client must rebuild the table using the new partitioning scheme prior to utilizing the Data Retention Manager Then you can update the values in the Data Retention Manager tables to reflect the new partitioning scheme Partition maintenance procedures remove old data from the database so that the database does not continue to grow until space is insufficient Daily weekly or monthly maintenance is necessary for those tables that have daily weekly and monthly pattitions respectively Partition maintenance 1 Copies information related to open orders from the oldest partitions to temp tables EXECUTION ORDR ORDR_EVENT ORDR_ STATE CHANG
43. and network metadata use the following procedure Metadata Note To avoid corrupting the Behavior Detection process never load scenarios while the process is running 1 Navigate to the following directory cd INSTALL DIR 2 db tools 2 Optional Edit the metadata configuration files that is senros c g dataset cfg and network cfg with identifying information for the scenarios datasets or netwotks that you want to load scnro ctlg id in the scnros cfg file and or scnro ctlg id inthe scnros cfg file Note Providing both scnro ctlg id and scnro id in the scnros cfg file allows finer granularity when loading scenarios You must separate values with a period per line data set id in the dataset cfg file network id in the network cfg file 282 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Scenario Migration Utility Chapter 8 Administrative Utilities 3 Copy the XML files you plan to load into the directory that the load dirname specifies in the install cfg file Figure 54 on page 207 4 Execute the sm_load sh script a Enter sm_load sh dataset to load dataset metadata b Enter sm_load sh scenario to load scenario metadata c Enter sm_load sh network to load network metadata Scenario Migration Best Practices Process Overview Best Practices Migrating scenarios from one environment to another requires a unified process
44. table name column column gt lt JOIN gt lt MULTIJOIN gt The following XML code provides an example lt ANALYSIS gt lt JOINS gt lt MULTIJOIN gt lt JOIN gt LEFT table ACCT column ACCT INTRL ID RIGHT table CUST ACCT column ACCT INTRL ID JOIN MULTIJOIN lt MULTIJOIN gt lt JOIN gt LEFT table ACCT column ACCT INTRL ID RIGHT table CUST ACCT column ACCT INTRL ID JOIN lt JOIN gt LEFT table CUST column CUST INTRL ID RIGHT table CUST ACCT column CUST INTRL ID JOIN MULTIJOIN lt JOINS gt lt ANALYSIS gt This XML code executes the following queries select count 1 from ACCT a CUST ACCT b where a ACCT INTRL ID b ACCT INTRL ID select count 1 from ACCT a CUST ACCT b CUST c where a ACCT INTRL ID b ACCT INTRL ID and c CUST INTRL ID b CUST INTRL ID Simple Join with Filter Restriction Adding a filter to the joins determines the join count between tables with a restriction A filter uses the table field operator and value attributes to set the restriction The operator is limited to the XML code operators in Table 101 on page 264 for more information The structure is organized in the same manner as a Simple Join with an added lt FILTER gt element The following code illustrates the structure lt MULTIJOIN gt lt JOIN gt lt LEFT table table name column column gt lt RIGHT table table name column c
45. 1 3 Administration Guide Index Post Watch List Miscellaneous 140 163 169 predecessors 134 Pre Watch List Miscellaneous 137 159 summary 143 151 155 164 170 Watch List 138 147 148 160 167 Watch List AML Brokerage 148 154 Ingestion Manager 8 51 install cfg file 198 262 Insurance 169 Intra Day Ingestion 73 J jobs monitoring 22 monitoring and diagnosing 20 performing 17 recover 24 restarting 20 21 starting 19 status codes 18 stopping 21 join counts analysis distinct column 269 simple join 268 jurisdiction about 35 geographical 35 KDD_JRSDCN table 35 organizational 35 usets 30 K KDD AUTO CLOSE ALERT table 178 KDD JOB TEMPLATE table 12 L LDAP 26 limitations report framework 315 Load Workflow 52 Loading Holidays 211 annual activities 211 Logging 299 configuration file 303 diagnostic 300 fatal 300 file sizes 307 location property 307 Location Property Values 307 message library 299 message template repository 300 monitoring 309 notice 300 properties 308 trace 300 warning 300 Logging levels Diagnostic 300 Fatal 300 Notice 300 Trace 300 Warning 300 logical name Report 314 Report Comments 315 Report Definition 314 Report Definition Parameters 314 Report Jurisdiction 315 Report Parameters 315 Report Template 314 Report Template Jurisdiction 314 Report Template Parameters 314 logs Calendar Manager Utility 236 disp
46. 325 Moving PROD Environment to UAT Appendix E Moving Oracle Financial Services System Environment 326 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Index A access control 34 metadata 34 preparation 33 active pages report management 313 Adding Entries directly in the Table using script 39 Adding Entries through Excel Upload 39 Administration Tools Web enabled 4 advanced alert creator configuration 174 rules 174 alert advanced alert creator configuration 174 auto close 172 auto close alert 182 correlation 172 creation 171 174 flagging duplicates 253 highlight generation 172 184 notification 172 scoting 172 suppression 172 suppression job 183 alert auto close 180 null value 180 rules 180 Alert Correlation 186 Alert Correlation Migration Utility logs 287 alert correlation rule metadata extracting 292 Alert Correlation Rule Migration Utility 287 alert correlation rule load 291 configuring 289 extracting metadata 292 loading metadata 294 scenario extraction 290 Alert Correlation Rules Migration Utility configuring 287 using 287 Alert Creation 171 172 174 186 alert creator 174 grouping algorithms 175 running multi match 174 running single match 174 Alert Management 3 Alert Purge Utility 195 215 217 configure 217 directory structure 215 executing 219 logs 216 precautions 216 processing 219 purge alerts 221
47. 8 Administrative Utilities Note Oracle Financial Services recommends that you view the output report using Microsoft Excel because this HTML file has specific HTML formatting for Excel Table 104 describes sections of the output report Table 104 Data Analysis Report Output Section Description Table Count Summary Contains the row count of each table in the configured data base excluding the KDD archive and temp tables Field Distribution Sum Groups by table the unique values for the identified fields mary Table and number of times each value occurs in the table This summary table appears only in the report if the analysis for Distinct Values for Fields of Interest and Its Count was con figured in the XML file In addition quotes enclose any val ues with padded spaces to identify spaces in the value Null Summary Count Groups by table the number of nulls present and values with Table padded spaces for the identified fields in each table This summary table only appears in the report if the analysis for Null and Padded Space Count has been configured in the XML file Referential Integrity Displays the join analysis the number of rows returned Table Summary between the joined tables and the table count for each table being joined This summary only appears in the report if the analysis for Join Counts has been configured in the XML file Query Results Displays the results of queries specified in t
48. Administration Guide AML Banking Informatica Workflows Chapter 5 Informatica Workflows Table 53 Watch List Workflows AML Banking Sequence Workflow Number Folder Number Workflow Name Predecessor 3 ORION Common 3000 w ph3000 Adjust WL WLS ORION Common 2180 3 ORION Common 3005 w ph3005 apply cust KYC risk 3 ORION Common 3030 w ph3030 truncate wls 4 ORION Common 3010 w ph3010 truncate Is 4 ORION Common 3020 w_ph3020_truncate_nms 4 ORION Common 3040 w ph3040 truncate wls2 4 ORION Common 3051 w ph3051 create addresses from fotps 4 ORION Common 3090 w ph3090 create external entities from fotps ORION Common 2190 ORION Common 3091 4 MLM Banking Com 3130 w ph3130 create client banks from fotps ORION Common 2190 mon 5 ORION Common 3070 w ph3070 load watch list staging table ORION Common 3000 ORION_Common 3030 5 ORION Common 3140 w ph3140 write fotps associations to Is ORION Common 3010 ORION Common 3051 ORION Common 3090 6 ORION Common 3100 w ph3100 load staging fuzzy matches ORION Common 3020 ORION Common 3070 ORION Common 3090 6 ORION Common 3180 w ph3180 write Is to link tables ORION Common 3110 ORION Common 3140 MLM Brokerage Common 3160 7 ORION Common 3150 w ph3150 load staging and validate watch list ORION Common 3040 ORION Common 3070 ORION Common 3090 ORION_Common 3 0
49. Administration Guide 167 Insurance Workflows Chapter 5 Informatica Workflows Table 67 Watch List Workflows Insurance Sequence Workflow Number Folder Number Workflow Name Predecessor 3 ORION_Common 3000 w ph3000 Adjust WL WLS ORION Common 2180 3 ORION Common 3005 w ph3005 apply cust KYC risk 3 ORION Common 3030 w ph3030 truncate wls 4 ORION Common 3010 w ph3010 truncate Is 4 ORION Common 3020 w ph3020 truncate nms 4 ORION Common 3040 w ph3040 truncate wls2 4 ORION Common 3051 w ph3051 create addresses from fotps 4 ORION Common 3061 w ph3061 create addresses from INS 4 ORION Common 3080 w ph3080 create external entities from INS ORION Common 2193 4 ORION Common 3090 w ph3090 create external entities from fotps ORION Common 2190 ORION Common 3091 5 ORION Common 3070 w ph3070 load watch list staging table ORION Common 3000 ORION Common 3030 5 ORION Common 3110 w ph3110 write INS associations to ls ORION Common 3061 ORION Common 3080 5 ORION Common 3140 w ph3140 write fotps associations to Is ORION Common 3010 ORION Common 3051 ORION Common 3090 6 ORION Common 3100 w ph3100 load staging fuzzy matches ORION Common 3020 ORION Common 3070 ORION_Common 3090 6 ORION Common 3180 w ph3180 write Is to link tables ORION Common 3110 ORION Common 3140 MLM Brokerage Common 31
50. Administrative Utilities Configuration files enable the utilities to share common resources such as database configuration directing output files and setting up logging activities Common resources include the following install cfg file Refer to the following section sstall cfg File on page 198 for mote information categories cfg file Refer to categories ofg File on page 208 for more information install cfg File Configuration information resides in the lt INSTALL_DIR gt database db_tools mantas_cfg install cfg configuration file The configuration file contains modifiable instructions for Oracle database drivers and provides information that each utility requires It also provides the user name and password that you need to connect to the database In this file you can modify values of specific utility parameters change the locations of output files and specify database details for extraction and data loading The install cfg file contains information unique to each utility and common configuration parameters headings in the file clearly identify a utility s parameters You can also modify the current logging configuration for example activate or deactivate particular logging levels and specify locations for logging entries Figure 54 which appears on the next several pages provides a sample install cfg file with common and utility specific information Logging information appears at the end of the file
51. Behavior Detection Platform 6 1 3 Administration Guide Auto Close Chapter 6 Post Processing Tasks 7 DMN CDis the Oracle Financial Services application product code e ThecLS ACTIVITY TYPE CD column specifies the activity type code of the closing action e The cMMNT TX column specifies an optional text comment Note The activity type code that CLS ACTCVY TYPE CD specifies must exist in the KDD ACTIVITY TYPE CD table and the KDD ACTIVITY TYPE CD Verify that the AUTO CLOSE FL is set to Y for this code to be valid Insert the needed rows into the KDD AUTO CLOSE ALERT table Sample Auto Closing Alert Rule You may want to close an alert when the match score is less than 75 and the status code is equal to NW New or the review is more than 30 days old If so follow the steps 1 Determine the ATTR ID for the columns to reference in the KDD REVIEW table SCORE has ATTR ID 113000022 STATUS has ATTR ID 113000008 AGE has ATTR ID 113000057 Formulate the operations The match score is less than 75 and the status code is equal to NW SCORE 75 AND STATUS NW Reviews more than thirty days old AGE gt 30 Determine an otder of precedence for the criteria For example to determine whether reviews are more than thirty days old assign AGE gt 30 a precedence of 1 and SCORE lt 75 AND STATUS NW a pre cedence of 2 Assign an operation ID to each operation within the operation s
52. Chapter 6 Post Processing Tasks Update Alert Financial Data Oracle Financial Services provides some enhanced data on alerts to support searching by alerts based on business data For example Trader focused alerts may be searched based on the security involved in the activity Update Alert Financial Data is the process that populates this information To update alert financial data run the following command from the INSTALL DIR database db tools bin directory upd kdd review fin sh batch id YYYYMMDD If batch id and the batch date lt YYYYMMDD gt are not provided the system derives this data for matches created in the current batch The log for this process is under the INSTALL DIR database db tools 1ogs directory The name of the file is run stored procedure log Alert Scoring Oracle Financial Services provides a mechanism to compute a score for alerts to provide an initial prioritization The score is an integer and will be bounded by a configurable minimum and maximum value This module has two different strategies for computing the alert s score All strategies are based on the score of the alert s matches The strategies are e Max Match Score The score of the alert equals the alert s highest scoring match e Average Match Score The score of the alert equals the average of its matches score Refer to the Oracle Financial Services Behavior Detection Platform Administration Tools User Guide Releas
53. DT Figure 8 Sample SQL Script for Loading KDD ORG Refer to Creating Jurisdiction in tbe Database on page 35 to perform the Excel Upload of organization The Excel template to be used is KDD ORG x1s Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 41 Mapping Users To Access Control Metadata Chapter 3 Security Configuration Mapping Users To Access Control Metadata An Administrator can map each user to Access Control Metadata and Security attributes which will control the user s access permissions The Security Attribute Administration can be accessed from the Administration menu Figure 9 Note Before proceeding with providing a user access through this UI all necessary data should be available in the appropriate database tables and the user must be created ORACLE Financial Services Enterprise Case Management Choo SEE User Pool Line Organization Parent Organization Own Case Flag Own Alert Flag Email Address Business Organization gt Organization 0 5 Expand All Remove Jurisdiction P Jurisdiction 0 9 Remove Business Domain Business Domain 2 9 Remove E Business Domain Code a b Scenario Class gt Scenario Class 0 5 Expand All Remove Case Type Subtype gt Case Type Subtype 0 5 Expand All Remove Correlation Rule gt Correlation Rule 0 Remove Administr
54. Data Before you use control data to test your Behavior Detection installation the following prerequisites must be fulfilled 1 The maximum lookback that control data considers is of 13 months which is for change in behavior scenarios Hence while creating control data ensure that it is spread over 22 different dates in 13 months 2 The current day according to control data is 20091210 3 Unless specified set the current date as 20091210 to generate alerts on control data before running Behavior Detection Platform Note For more information about control data on your site contact your Oracle Financial Services Administrator 128 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Use of Control Data Chapter 4 Data Ingestion Control Data Ingestion Control data uses a specific set of dates to ensure that all the scenarios are tested using this data The maximum lookback that control data considers is of 13 months which is for change in behavior scenarios The control data is spread over 22 different dates in 13 months The dates YYYYMMDD format being used by control data are Table 43 Dates used by Control Data 20081231 20091124 20090130 20091125 20090227 20091126 20090331 20091127 20090430 20091130 20090529 20091201 20090630 20091202 20090731 20091203 20090831 20091204 20090930 20091208 20091030 20091209 20091123 20091210 O
55. Excel Upload for Business Domain The Excel template to be used is KDD BUS DMN xl1s Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 37 About Configuring Access Control Metadata Chapter 3 Security Configuration Creating Case Type Subtype If your firm has implemented Oracle Financial Services Enterprise Case Management you will need to establish access permissions associated with the available Case Types and Subtypes Case Type Subtype is used for data access controls similar to business domain but have a different objective The case type subtype can be used to identify records of different case types or to provide more granular restrictions to data such as case data The following tables are involved in the display of the Case type Subtype SubClass1 and SubClass2 in the Case Management UI and are specific to the Case Management implementation KDD_CASE_TYPE_SUBTYPE Each record in the Case Type Subtype table represents a case type subtype available in the Oracle Financial Services Case Management system Cases are logically grouped to a certain type and subtypes based on their behavior of interest and purpose of investigation like AML Fraud etc When generated a case should be mandatorily assigned to one of the case types for further investigation For a case type subtype is may or may not exist KDD_SUBCLASS1 Each record in the Case Subclass 1 table represents a subclass based on whi
56. Financial Services Job Protocol Performing Dispatcher Tasks Performing Job Tasks Clearing Out the System Logs Recovering Jobs from a System Crash About the Oracle Financial Services Job Protocol The system initiates all jobs by using a standard operational protocol that utilizes each job s metadata which resides in a standard set of database tables Oracle Financial Services Job Protocol processes include the following dispatcher Polls the job metadata for new jobs that are ready for execution This daemon process starts a mantas process for each new job mantas Creates a new job entry based on a template for the job that has the specific parameters for this execution of the job that is it clones a new job Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 11 About the Oracle Financial Services Job Protocol Chapter 2 Behavior Detection Jobs As the Oracle Financial Services administrator you invoke the dispatcher and mantas processes by running the shell scripts in Table 3 Table 3 Shell Scripts that Call mantas Processes Process Description start_mantas sh Starts all jobs This script invokes the cloner and mantas processes This is the integration point for a third party scheduling tool such as Maestro or AutoSys start chkdisp sh Calls on the check dispatch sh script to ensure that the dispatcher runs stop chkdisp sh Stops the dispatcher process restart
57. INSTALL DIR database db tools mantas cfg install cfg configuration file contains modifiable inputs that you use to tun the utility Refer to install ofg File on page 198 for more information This section contains the following topics e Directory Structure Refer to Directory Structure on page 236 for more information e Logs Refer to Logs on page 236 for more information e Calendar Information Refer to Calendar Information on page 237 for more information e Using the Calendar Manager Utility Refer to Using the Calendar Manager Utility on page 237 for more information Table 92 provides the directory structure for the Calendar Manager Utility in INSTALL DIR database db tools Table 92 Calendar Manager Utility Directory Structure Directory Description bin Contains executable files including the shell script set mantas date sh lib Includes required class files in jar format mantas cfg Contains configuration files for example install cfg and categories cfg in which you can configure properties and logging attributes log Keeps the calendar_manager 1og log file that the utility generates during execution As the utility updates the calendars in the system it generates a log that it enters in the INSTALL DIR database db tools logs calendar manager log file the logging process time stamps all entries The log file contains relevant information such as statu
58. Manager activities Table 97 describes the table s contents Table 97 BUSINESS KDD DR MAINT OPRTN Table Contents Column Name Description PROC ID Identifies the sequence ID for the operation to perform ACTN TYPE CD Indicates the activity that the utility is to perform on the table A Analyze RI Rebuild Indexes e P Partition e RV Recompile Views OWNER TABLE NM Identifies an owner or user of the utility Identifies a database table Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 249 Data Retention Manager Chapter 7 Batch Processing Utilities Table 97 BUSINESS KDD DR MAINT OPRTN Table Contents Continued Column Name Description PARTN TYPE CD Indicates the partition type e D Daily e W Weekly e M Monthly e X Mixed Date TOTAL PARTN CT Specifies the total number of partitions to be created including the current partition For example for a daily partitioning scheme of four previous days and the current day the value of this field is five 5 BUFFER PARTN CT Specifies the number of buffer partitions the utility is to maintain excluding the cur rent partition For example a two day buffer has a value of two 2 CNSTR ACTN FL Determines whether to enable or disable constraints on the table during processing WEEKLY ADD FL Indicates whether daily partitions are added for a week at a time If
59. Note Oracle Financial Services Behavior Detection supports only one method of managing trusted pairs per installation Clients may elect to create and manage trusted pairs through the loading of trusted pairs via a DIS file or utilize the Behavior Detection user interface for creation and management of trusted pairs However both the methods should not be utilized concurrently Copying Thomson Reference Files Behavior Detection uses the Thomson Financial Publishing US and Non US institution lists These lists are used to cross reference bank identifiers of different types and different resolution for example FedWire 8 digit BIC and 11 digit BIC The system matches banks that appear in transaction data to the Thomson data and generate the Institution Master table Table 41 lists reference files and their sources If the reference files are not available Behavior Detection derives only Institution Master records for accounts of the Firm that are part of a bank role Thomson Financial Publishing updates these files monthly Behavior Detection must obtain updates directly from Thomson Financial Publishing Table 41 Reference Files Short Name Description Source Processing Guidelines US Banks Non US Banks List of all financial institutions in the United States List of all financial institutions outside the United States Thomson Financial Publishing Electronic Payments File rtsubase txt Thomson Fin
60. P NETACENCU NT CUACTAXID TMP NETACENCU NT JRNL P NETACENCU NT WIREACBENE TMP NETACENCU NT WIREACORIG Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 257 Truncate Manager Chapter 7 Batch Processing Utilities CST Losses Truncate Manager Logs DB TOOLS refresh temp table sh TMP NETACENCU NT WIRETRXN TMP NETACENCU NT WIRETRXN VW 2 Execute the link analysis network generation job The product job template ID is 114698120 3 Execute the scenario job The product job template ID is 117350084 To populate the temporary tables for CST LOSSES scenario follow the steps 1 Execute this refresh temporary table process DB TOOLS refresh temp table sh VWCST_1OSSES AC ASM TMP VWCST lOSSES AC ASM 2 Execute the link analysis network generation job 3 Execute the scenario job The Data Ingestion subsystem calls the run_truncate_manager sh script to truncate tables that require complete replacement of their data The log category TRUNCATE_MANAGER location property in the INSTALL DIR database db tools mantas cfg categories cfg file controls logging for this utility The system writes log information for this process to the following location INSTALL DIR database db tools logs truncate manager log Using the Truncate Manager The run truncate manager sh script takes the table name as an argument the table must exist in the BUSINESS
61. PTA tasks must run in this order 1 Load Employee User Groups 2 Request Assignment After both scripts have been successfully executed upon initialization of the AA PTA application both scripts can be run independently of each other and how frequent each script is run is determined by the Oracle client 194 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide CHAPTER 7 Batch Processing Utilities Oracle Financial Services Behavior Detection Platform provides utilities that enable you to set up and modify a selection of batch related database processes The chapter focuses on the following topics About Administrative Utilities About Annual Activities Alert Purge Utility Batch Control Utility Batch Export Utility Calendar Manager Utility Data Retention Manager Database Statistics Management Flag Duplicate Alerts Utility Notification Refreshing Temporary Tables Truncate Manager ETL Process fot Threshold Analyzer Utility Process to Deactivate Expired Alert Suppression Rules About Administrative Utilities Behavior Detection database utilities enable you to configure and perform batch related system pre processing and post processing activities Alert Purge Provides the capability to remove erroneously generated matches alerts and activities Refer to Avert Purge Utility on page 215 for more information Batch Control Manages the start and termination of a batch process from D
62. Pages Desktop Prerequisites for Integrating Report Framework Before you use Report Framewotk the following prerequisites must be fulfilled 1 Altio s p fservlet war file must be present on Altio Presentation Server 2 Report Framework must be installed on the same Altio Presentation Server as the client application 3 A PDF template with form fields to act as placeholders must be present in the templates directory The PDF template can be created using any PDF Writer 4 The client App should contain a function called generatePDF that invokes an HTTP Service This service makes a call to the PDF Generation servlet This Altio function determines the data it requires to archive as a PDF 5 Configure the following parameters in the client app to enable you to connect to the Report Framework pdfservlet war location e PDF Template location Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 313 Report Management Appendix C Report Management Table 119 Metadata Creation for Report Framework 6 Set the parameters in the client app to be consider as default in the KDD REP XXX tables The set parameters are disabled when the client app is launched from the Report Management Console Refer to Table 119 for more information on Metadata Table 119 Refer to the Metadata creation for Report Framework Logical Name Report Definition Table Name KDD REPORT DEFN Table
63. Pre Trade Approval Roles and User Groups Role Group Name User group Code Employee Employee User Group CREMPLOYEEUG Control Room Analyst Control Room Analyst User CRANALYSTUG Group Control Room Supervisor Control Room Supervisor User CRSUPVISRUG IP Manager Group IP Manager User Group CRIPMANAGERUG IP Manager Supervisor IP Manager Supervisor User CRIPMGRSUPVSRUG Group While it is possible for a user to have more than one role within an application care must be taken with the combination of roles Specifically a user should be assigned only one role from among the following AM Analyst I AM Analyst II AM Analyst III AM Supervisor AM Executive AM Internal and External Auditors A user with one of these AM users can otherwise be assigned additional role s from among AM Scenario Group AM Administrator and a Case Management role Similarly a user with a Case Management role should be assigned only one role from among the following Case Analyst Case Analyst2 Case Supervisor Case Executive Case Internal and External Auditors Case Viewer Case Initiator A user with one of these Case Management roles can be assigned additional roles from among the AM roles or the Case Administrator role Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 29 Setting up a User Chapter 3 Security Configuration Mapping a User toa Single User Group M
64. Pre Watch List Workflows AML Banking Pre Watch List Informatica Miscellaneous workflows facilitate the application of customer supplied measures of risk to corresponding entities transactions and instructions Figure 35 illustrates Informatica Pre Watch List Miscellaneous workflows for AML Banking Sequence 1 2005 2140 2180 2600 Note Processing can occur simultaneously within each sequence Workflows with predecessors 2190 3091 must wait for predecessor results before workflow processing can begin Sequence 2 Figure 35 Independent Pre Watch List Workflows AML Banking The system completes these workflows in two sequences processing of the workflows does not require data from external predecessors These workflows may proceed simultaneously in all sequences as soon as processing of any predecessors has completed Table 52 describes the Informatica Miscellaneous Independent Pre Watch List workflows in Figure 35 Table 52 Independent Pre Watch List Workflows AML Banking Sequence Number Folder Workflow Number Workflow Name Predecessor 1 ORION Common 2005 w p2005 set cd pd dates 1 ORION Common 2140 w ph2140 pass thru process 1 ORION Common 2180 w ph2180 instn identification 1 ORION Common 2600 w ph2600 loan smry mnth ORION Common 2190 w ph2190 fotps instn processing ORION Common 2180 ORION Common 3091 w ph3091 create PartyNmEE from fotps Oracl
65. Software s Short Introduction to log4j guide http log ging apache org log4j docs manual html for more details about the log message format log message library To be specified at installation Identifies the full path and filename of the message library log max size 2000000000 Determines the maximum size in bytes of a log file before the system creates a new log file For more information Refer to Log File Sizes on page 307 for more informa tion location log categories file path log category catgory name To be specified at installation Contains routing information for message libraries for this category For more infor mation Refer to Logging Location Property Values on page 307 for more information Identifies the full path to the categories cfg file log category name lt severity gt location Contains routing information for message libraries with the given severity for the given category For more information Refer to Logging Location Property Values on page 307 for more information log4j config file To be specified at installation Specifies the full path to the external log4j configuration file log default location log mantaslog location Contains routing information for message libraries for this category for which there is no location previously specified Contains routing information for message libraries for this category fo
66. Workflow Name Predecessor 13 ORION Common 2172 w ph2172 update jurisdiction in ag 13 ORION Common 2200 w ph2200 build trxn tables from fotps ORION Common 3051 ORION Common 3210 14 ORION Common 2220 w ph2220 update fot reversals ORION Common 2200 20 ORION Common 3501 w ph3501 Exp and Risk Review TP ORION Common 2200 21 ORION Common 3502 w ph3502 Flag Trusted Trxn ORION Common 3501 Informatica Summary Workflows Fraud Detection Informatica Summary workflows maintain monthly aggregations of customer activity Figure 48 illustrates Informatica Summary workflows for Fraud Detection vedac seure 2005 3210 0200 0371 0471 Misc Watch List Misc Misc Misc Sequence 16 0180 0190 Sequence 17 0270 Sequence 18 0310 Sequence 19 Note Processing can occur 0470 0630 0640 0750 Sequence 20 0280 simultaneously within each sequence Sequence 21 Workflows with predecessors 0760 must wait for predecessor results before workflow processing can begin Figure 48 Summary Workflows Fraud Detection The application completes these workflows in five sequences processing of some workflows requires data from Watch List and Miscellaneous predecessor workflows refer to Informatica Miscellaneous Optional Workflows Fraud Detection on page 158 and Informatica Watch List Workflows Fraud Detection on page 160 for more information 164 Oracle Financial Services Behavior Detection Platform
67. You must first provide the user with access privileges so the user can perform activities throughout various functional areas in Behavior Detection and Account Approval Pre Trade Approval This enables the user to access at least one of each of the following Jurisdiction Scope of activity monitoring for example Geographical Jurisdiction or Legal entity Refer to Creating Jurisdiction in the Database on page 35 for more information Business Domain Operational line of business Refer to Creating Business Domain in the Database on page 36 for more information Scenario Group Grouping of scenarios to control user access to scenarios Role Permissions or authorizations assigned to a user Organization User group to which a user belongs Some data types such as Scenario Group Role Business Domain Case Type and Case Subtype which compose the user security configuration are predefined with sample values which are available through the installer Clients can change or add new values for these data types with the exception of User Role based on specific requirements The following section explains how to add or modify these data types 34 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide About Configuring Access Control Metadata Chapter 3 Security Configuration Creating Jurisdiction in the Database Creating Jurisdiction in the Database through Scripts Behavior Detection and
68. a Batch Process To End a Batch Manually Processing for End Batch When a batch ends as part of an automated process the utility retrieves the batch name and other information from the KDD PRCSNG BATCH table Refer to Table 83 on page 226 To stop a batch process manually follow the steps 1 2 Verify that the Behavior Detection database is operational tnsping database instance name gt Verify that the INSTALL DIR database db tools mantas cfg install cfg configuration file contains the correct source database connection information Access the directory where the shell script resides cd INSTALL DIR gt database db tools bin Start the batch shell script end mantas batch sh If you enter an invalid batch name the utility terminates and logs a message that desctibes the error The error message appears on the console only if you have output to the console enabled in the INSTALL DIR database db tools mantas cfg categories cfg configuration file After establishing the required Java environment and initiating various Java processing activities the Batch Control Utility does the following 1 2 Determines whether a batch is running by querying the KDD PRCSNG BATCH CONTROL table Refer to Table 87 on page 228 Records information about the batch in the KDD_PRCSNG_BATCH_ HIST table Refer to Table 88 on page 228 This table contains a history of all batches that appeat by start
69. alert correlation rule e The attribute value for jpe attribute in the XML tag lt Case gt should exist in the CASE TYPE CD column of the CASE schema table KDD CASE TYPE SUBTYPE e The attribute value for subfpe attribute in the XML tag Case should exist in the CASE SUB TYPE CD column of the CASE schema table KDD CASE TYPE SUBTYPE 294 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Alert Correlation Rule Migration Utility Chapter 8 Administrative Utilities e The attribute value for lt subClassTagLeveli gt should exist in the CASE SUB CLASS LVL1 CD column of the CASE schema table KDD_SUBCLASS1 e The attribute value for lt subClassTagLevel2 gt should exist in the CASE SUB CLASS LVL2 CD column of the CASE schema table KDD SUBCLASS2 e The CDATA section of the XML tag lt AlertAttrOperations gt is validated as follows The valid operations should be one of BOTH TO and FROM The valid operators can be gt lt gt IN NOT IN AND and OR The TO and FROM alert operations can be used only to compare alert attribute values to each other and not to a literal The FROM alert operation should always precede the TO alert operation The BOTH alert operator must be used to compare alert attribute values to a literal The expression can be nested to any arbitrary length provided that it confirms to a general syntax Operand Operator Operand Logical_
70. alert or case ID Attachments associated with the alerts and cases being exported are copied to the child directoty corresponding to the alert or case to which they are associated The exported data includes XML metadata for easier searching of the archived documents The XML file name uses the following file formats Alert AlertID info xml Case CaseID info xml Note Alert metadata xsd lt alert mgmt gt WEB INF classes conf ui_config dtd mantasAlertExportD ata xsd Alert Metadata System Name Hard coded to Oracle Financial Services Behavior Detection PDF File name Element ExportDate System Date Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 233 Batch Export Utility Chapter 7 Batch Processing Utilities Table 90 Alert Metadata Continued Alert Metadata Element AlertID KDD REVIEW REVIEW ID Element CreationDate KDD REVIEW CREAT TS Element CurrentStatus KDD REVIEW STATUS STATUS CD Element FocalEntityDisplayID KDD REVIEW FOCAL NTITY DSPLY ID Element FocalEntityName KDD REVIEW FOCAL NTITY DSPLY NM Element FocusTypeCode KDD CENTRICITY CNTRY TYPE CD Element ScenarioDisplayName KDD REVIEW SCNRO D
71. an analysis is possible if the lt CONSTRAINT gt element identifies the filter The following code illustrates the structure for using a filter TABLE name table name gt lt get distinct value for one column gt VALUES COLUMN field column name gt lt Constraint feature is optional May contain one or more constraints gt CONSTRAINT field column name operator operator value filter value gt lt VALUES gt lt get distinct value for many columns gt lt VALUES gt lt COLUMN field column name gt COLUMN field column name gt lt Constraint feature is optional May contain one or more constraints gt CONSTRAINT field column name operator operator value filter value gt lt VALUES gt lt TABLE gt The following XML code illustrates use of a filter lt ANALYSIS gt lt TABLES gt lt TABLE name ACCT gt VALUES COLUMN field ACCT TYPE1 CD COLUMN field ACCT TYPE2 CD lt VALUES gt lt TABLE gt lt TABLE name CUST gt lt VALUES gt lt COLUMN field CUST_TYPE_CD gt Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 265 Data Analysis Tool Chapter 8 Administrative Utilities CONSTRAINT field DATA DUMP DT operator EQ value 15 NOV 2006 gt lt VALUES gt lt TABLE gt lt TABLES gt lt ANALYSIS gt This XML code executes the following queries select ACCT TYPE1 CD ACCT TYP
72. and KDD CAL HOLIDAY tables the utility creates entries in KDD_CAL 238 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Calendar Manager Utility Chapter 7 Batch Processing Utilities Table 93 provides the contents of the KDD_CAL table Table 93 KDD_CAL Table Contents Column Name Description CLNDR_NM Specific calendar name CLNDR_DT Date in the range between the lookback and lookforward periods CLNDR_DAY_AGE Number of calendar days ahead or behind the provided date The provided date has age 0 the day before is 1 the day after is 1 For example if a specified date is 20061129 the CLNDR DAY AGE of 20061128 1 and 20061130 1 BUS DAY FL Flag that indicates whether the specified date is a valid business day set the flag to Y Set this flag to N if the DAY OF WK column contains an entry that appears as a valid non business day in the KDD CAL WKLY OFF table or a valid holiday in KDD CAL HOLIDAY BUS DAY AGE Number of business days ahead or behind the provided date If BUS DAY FL is N BUS DAY AGE receives the value of the previous day s BUS DAY AGE BUS DAY TYPE CD Indicates the type of business day e N Normal e C Closed e S Shortened DAY OF WK Value that represents the day of the week Sunday 1 Monday 2 Tuesday 3 Saturday 7 SESSN OPN TM Indicates the opening time of the trading se
73. are 10 eads used to initialize the SSA PERSONAL NAME and SSA BUSINESS NAME tables that contain Informatica Identity Resolution keys corresponding to names in the WATCH LIST table data Subdirectory The data subdirectory within the data ingest directory contains additional subdirectories for organizing Market data files and Oracle client data files The system creates these files during the preprocessing transformation and data loading stages of the ingestion process The Market data and Oracle Financial Services client data files appear in subdirectories that are indicative of the processing stages or workflow steps that the Data Ingestion subsystem components perform The following sections describe the contents of each subdirectory and the components that read or write to each subdirectory data errors The errors subdirectory within the data subdirectory stores error files that Data Subdirectory Ingestion subsystem components create or move upon detection of errors during file processing The system places error files in subdirectories within the errors subdirectory These error file subdirectories are name based on the processing date for the files that they contain The date has the format YYYYMMDD where vvvY is the four digit year MM is the two digit month and DD is the two digit day The files in the errors subdirectory have the same name as the file in which the error was detected However the component that identified the errors a
74. backup files for the various Data Ingestion components refer to data backup Subdirectory on page 105 for more information data errors Contains error files for various Data Ingestion components refer to data errors Subdirectory on page 103 for more information data firm Contains Oracle client data files that Data Ingestion components write refer to data firm Subdirectory on page 105 for more information data market inbox informatica Contains market data files that Data Ingestion components write refer to data market Subdirectory on page 104 for more information Contains data files that the Oracle client provides refer to nbox Subdirectory on page 106 for more information Identifies the root directory for Informatica components and directories As part of the installation process the system moves files for Informatica ingestion components to appropriate directories For more information about installation refer to the Oracle Financial Services Behavior Detection Platform Installation Guide Release 6 1 3 jars Contains the Java Archive JAR files to run Java Data Ingestion components implemented in Java refer to bin Subdirectory on page 82 for more information logs Contains log files that Data Ingestion components write refer to logs Subdirectory on page 107 for more information Scripts Contains all the shell scripts for running Data Ingestion compon
75. batch Obtain status of a running batch run get mantas batch sh End a batch manually As Needed Alert Purge Scenario Migration Override Dataset Get Dataset Query Monthly Tasks Data Retention Configure monthly database partitioning run Remove erroneously run drm utility sh generated matches with alerts and activities drm partition type M run purge alert sh Extract dataset query and used as input run get dataset query sh Figure 53 Managing Database Activities with Utilities As Figure 53 illustrates daily tasks are initially dependent on the annual tasks that you perform such as obtaining holiday and weekly off days from an Oracle Financial Services client Daily tasks can include updating Behavior Detection calendars and managing batch processes You may need to configure data partitioning on a daily weekly or monthly basis Migrate scenarios data sets networks and associated meta data run sm extract sh sm load sh Permit dataset customizations populate dataset lt DATASET_ID gt txt Tasks that you perform when needed can include deleting extraneous or invalid matches and alerts or migrating scenarios and other information from one environment to another for example from test to production Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 197 About Administrative Utilities Chapter 7 Batch Processing Utilities Common Resources for
76. case the resulting type subtype subclass tag level 1 and subclass tag level 2 of the case are determined by the Ape subtype subClassTagLevell and subClassTagLevel2 attributes of the Case element The focus of the case is determined by using the ordered list of preferred business entity types defined in the FocusTypePref element In the example above if the alerts involved in the associated correlation are correlated to the same CUSTOMER then CUSTOMER would become the focus of the case If not and if they are correlated to the same ACCOUNT ACCOUNT would become the focus of the case If not the correlation will not be promoted to a case Running the Alert Correlation job will execute only those correlation rules that are designated as Active Rules that are designated as Inactive will be ignored and not executed To deactivate an active correlation rule the correlation rule metadata must be modified to change KDD_CORR_RULE STATUS_CD from a value of ACT to NULL To activate an inactive rule modify KDD CORR RULE STATUS CD from a value of NULL to ACT Changes made to the metadata are effective immediately and will be utilized the next time alert correlation is run Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 191 Account Approval Pre Trade Approval Tasks Chapter 6 Post Processing Tasks Sample Alert Oracle Financial Services Behavior Detection delivers two sample alert correlation Correlation Rules
77. contains specifies the full path to a subsystem s message library file Logging Levels Table 116 outlines the logging levels that the Common Logging component supports Table 116 Logging Levels Severity Log Level Usage Fatal Irrecoverable program process and thread errors that cause the application to terminate Warning Recoverable errors that may still enable the application to con tinue running but should be investigated for example failed user sessions or missing data fields Notice default High level informational messaging that highlights progress of an application for example startup and shutdown of a process or session or user login and logout Diagnostic Fine grained diagnostic errors used for viewing processing status performance statistics SQL statements etc Trace Diagnostic errors use only for debugging purposes as this level enables all logging levels and may impact performance The configuration file specifies enabling of priorities in a hierarchical fashion That is if Diagnostic is active the system enables the Notice Warning and Fatal levels 300 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Logging Message Libraries Appendix A Logging Logging Message Libraries Administration Tools Database Scenario Manager Services Some Behavior Detection subsystems produce log output files in default locations The following
78. date Refer to Novification on page 254 for more information Refreshing Temporary Tables Refreshes temporary tables that the behavior detection process uses and estimates statistics for the newly populated tables Refer to Refreshing Temporary Tables on page 255 for more information Truncate Manager Truncates tables that require complete replacement of their data Refer to Truncate Manager on page 258 for more information Figure 53 illustrates the frequency with which you use these batch related database utilities when managing activities daily weekly monthly annually or as needed 196 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide About Administrative Utilities Chapter 7 Batch Processing Utilities Annual Tasks Holidays from Business Client Weekly Offs from Business Client Get holidays from business client late KDD CAL HOLIDAY Get weekly off days from business client populate KDD CAL WKLY OFF Weekly Tasks Data Retention Configure weekly database partitioning run run drm utility sh with drm partition type W Daily Tasks Calendar Update Batch Control Data Retention DRM Update calendars in system run set mantas date sh Start a batch manually run start mantas batch sh Maintain daily database partitioning DRM run run drm utility sh Gather statistics during Flag duplicate alerts run flag duplicate alerts sh the
79. e specified below the above synchronous properties must be set accordingly for the levels for which this detailed information is desired Also note that this type of detailed information can only be obtained for Java code log format d t p m n Specify the full path and file name of the message library log message library GWORKFLOW LOG MESSAGE LIB FILE Specify the full path to the categories cfg file log categories file path GWORKFLOW LOG CATEGORY PATHOG Multiple locations can be listed for each property using a comma delimiter log category TEST CATEGORY location console mantaslog log category TEST CATEGORY 2 location console users jsmith logs mylog log Continued on next page Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 305 Logging Configuration File Appendix A Logging Continued from previous page Specify where messages of a specific severity and category should be logged to The valid values are the same number as for category Multiple locations can be listed for each property using a comma delimiter If an entry for a severity is not listed here the message is logged to the location specified for the category number by the above property and if that does mot exist it is logged to the default location configured below HH HH log TEST CATEGORY warning location syslog log TEST CATEGORY fatal location userGdomain com log TES
80. e Ee 308 Monitoring Lop Filesiiss E 309 APPENDIX B Oracle Financial Services Software Updates 311 Behavior Detection Platform Software Updates Hotfix eene tnnt 311 Hotfix Effect on Customization DE V 312 User Tiiterta Ce MH 312 SCenatioss un Goin dide seo oto id dt dotis ehahebuununtaudi adi hi hia taies 312 APPENDIX C Report Management eeeeeeeeei eiiis eee en nnn nana nnn n nnns 313 R pott Maribement ee eet tete e LEE LED Lei LAM LL cies eet 313 Prerequisites for Integrating Report Framework sse nete 313 Report Framework T AmlalOHs uii prt torii eir ertet tera ipie etiain etie it edet 315 APPENDIX D Informatica Workflow Details eoe ssessss 317 Tnforimatica Workflows C 317 APPENDIX E Moving Oracle Financial Services System Environment 325 Moving PROD Environment to UAT iicet retener eo er eee k este EKS eere hoses repere veio rende 325 VINO ON qe 327 xiv Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide List of Figures Figure 1 Oracle Financial Services Architecture Over View esses 2 Figure 2 Oracle Financial Services Architecture Deployment View 5 Figure 3 Oracle Financial Services Architecture Security View esses 6 Figure 4 Oracle Financial Services Architecture Oracl
81. e Informatica Balances and Positions Workflows Broker Compliance e Informatica Miscellaneous Post Watch List Workflows Broker Compliance Informatica Miscellaneous Optional Workflows Broker Compliance Optional Miscellaneous workflows perform processing in support of workflows in multiple functional areas Figure 39 illustrates Optional Informatica Miscellaneous workflows for Broker Compliance Sequence 0 0010 0040 0050 0060 Note Processing can occur simultaneously within each sequence 0130 0070 0110 0120 Workflows with predecessors must wait for predecessor results before workflow processing can begin Figure 39 Optional Workflows Broker Compliance The application completes these workflows in one sequence processing of the workflows does not require data from external predecessors These workflows may proceed simultaneously in all sequences as soon as processing of any predecessors has completed Table 56 describes the Informatica Miscellaneous Optional workflows in Figure 39 Table 56 Optional Workflows Broker Compliance Sequence Number Folder Workflow Number Workflow Name Predecessor 0 ORION_Production 0010 w_p0010_dump_asm_to_file 0 BSM_Production 0040 w_p0040_dump_masm_to_file 0 BSM_Production 0050 w_p0050_dump_nmsm_to_file 0 BSM_Production 0060 w_p0060_dump_rrsm_to_file 0 ORION_Production 0070 w_p0070_reloa
82. e Nindicates that the firm calculates this field and provides the value during ingestion S SRISK ZONE 1 LOWER Defines the risk zones for Trusted Pairs S SRISK ZONE 1 UPPER functionality Each risk zone has a lower S SRISK ZONE 2 LOWER and upper bound The ranges of risk values S SRISK ZONE 2 UPPER within each zone are configurable but the SSRISK ZONE 3 LOWER number of risk zones shall remain at 4 If S SRISK ZONE 3 UPPER an implementation chooses not to use all SSRISK ZONE 4 LOWER Risk Zones then they can disable them by S SRISK ZONE 4 UPPER setting the risk ranges out of bounds For example Risk Zone 1 and Risk Zone 2 may have a lower and upper value of 0 Ingestion uses these risk zones to determine whether a party s effective risk has increased by enough points to move it to a higher risk zone and accordingly creates a new version of the trusted pair record with a status of Risk Esc Rec Cancel RRC 76 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Alternatives to Standard Data Ingestion Practices Chapter 4 Data Ingestion Note On AIX and Solaris the Informatica Ingestion process occasionally returns an invalid SQLException java sql SQLException ORA 01850 hour must be between 0 and 23 when running the s_m_p0403_load_wl session The S m p0403 load wl session runs successfully despite the reported error This etror should be ignored Alternative
83. eissira an aa aaa e aan a scans te eie nte 104 d ta baitk p Subdirector tM 105 data f firm S nodiri Sassari ihc a N A nab oe ttn Rn ok doa EE AEAEE 105 inbox Subditectoty uu 3t ete cett P ove ii Ra A ee i a a aS 106 tibspool S bditectoty M 106 LD SPOOL n S Ubre torja aiaa E pe d che bite bte anello bi ERU amd RU te p ade ed 106 Hbspooli backup Subdirectora gis ca eho opin eB sh Neh EEKE KE VCA E RENIER 107 ocean ER 107 Statt p and Shbu tdostis cite te tpi DEED Pb HERI PH UHR E GE e ERN RUIN Ho EE NR iat 108 Backup Server CofA NON orte iecore enirn poenitet teen etr e RARIUS FREUE 108 RECOVERY Je eoi e e e ERR E eu es ie o Su De rete E s Meuse Suns eee dup 108 Distributed Processing Configuration sse nete tete tenete tenente tenete nis 109 Data Rejection During IneestiOn scienter ect teet te Dee e etie LER c HE sein Coco ITI Rar Vel ROLE EEES 109 Rejection During the Preprocessirig Stage uie dotes nere br eger o ro ines iuge e triones eain 109 Duta LW m sVatebuisuelgted ttued 110 Missing Dataa aon RR ERR e oe Dae c eite e d bed Rated vei e 110 ODIT tr p iS 110 Domain Valts ri see n a ente eee bene Bn Chee 110 Rejection During the Transformation Stage seen tete tentent teens 110 oot EPOR Se sack iN ast iS ec decl cn dox ost pde icit ec d a pte ee 111 OD AY Da M n 111 Rejectio
84. enables proper display or analysis of these results may be required Note The Match Augmentation process is no longer explicitly run as a separate 8 8 job It is automatically executed at the end of each scenario run Match Scoring Computes a ranking for scenario matches indicating a degree of risk associated with the detected event or behavior Refer to Match Scoring on page 173 for more information Alert Creation Packages the scenario matches as units of work that is alerts potentially grouping similar matches together for disposition by end users Refer to Update Alert Financial Data on page 176 for more information Update Alert Financial Data Records additional data for alerts such as the related Investment Advisor or Security involved in the alert Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 171 About Post Processing Chapter 6 Post Processing Tasks Note Alert Scoring Ranks the alerts including each match within the alerts to indicate the degree of risk associated with the detected event or behavior Refer to Alert Scoring on page 176 for more information Assignment Determines the user or group of users responsible for handling each alert or case Refer to Assignment on page 177 for more information Auto Close optional Closes alerts that are of a lower priority to the business Refer to Auwto Close on page 178 for more information Automatic Alert Su
85. et ae vr repre rentes irte else dre ettet ote piens 254 Bath Based cessa o t RERO d i eode e a ede tee cates Ut ee Ee E e 254 hOoucclrdivocadMbii E 255 TiO GS cise thet sec no s te nr dette aia aa au da Pepi eet eee tede io tevendtoc a e ETENEE 255 Using Retreshine Temporary Tables oo otto tosta i ade o e e ade e use rone 255 Populate Temporary Lables tot Scenarios erint ee epe editi bap EE EREE 255 IMEEliddenRelationshebs dliN S T ertt rro d ea deno dao queitiqee diese e dede 255 VET IN ewr k Of AEN JA gu eibtetnae 256 FERN emok OAE JAC ssa ete er tt rre eer a ebrei PE ets 257 ANE 258 TtuncateManapet i i eene ute tete eerte tee rir Mes PRIN UE he ERI Ep eek PE EEEE ERE ebd ES EAE EES OSEN 258 I NE 258 sing the Truncate MA UDUMUI SC 258 ETL Process for Threshold Analyzer Utility esses atiii tenente entente entente 259 Process to Deactivate Expired Alert Suppression Rules sees 260 CHAPTER 8 Adminis trative Utililies cbe nbn 261 About Administrative Otilitiess csssssssssssssssssctscsssccusestessseschoesaievestovacsesstvsvascsussstsasensussbesasetveanascencusessaseveastbeasencusesestotoes 261 Common Resources for Administrative Utilities sse tentent 261 Data An
86. existing correlation of a correlation that has previously been identified ExtendFlag defines whether this correlation rule can result in extending an existing correlation If this is set to FALSE do not extend then a new correlation is created when this condition is identified If it is set to TRUE then the existing correlation is added to unless it has already been promoted to a case that has a status identified in the Casestatus tags of NonExtendableCaseStatuses Actions optional Once correlations ate discovered multiple types of actions can be taken on the correlation These actions and their associated parameters are defined in between the Actions tags The current set of possible actions include scoring the correlation and promoting the correlation to a case Scoring optional The strategy attribute defines whether the correlation score should be derived from the max of the associated alert scores MAX SCORE or the average of the associated alert scores AVERAGE SCORE The zucS rategy attribute provides the option of defining a fixed score to be added to the 190 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Alert Correlation Chapter 6 Post Processing Tasks Activating or Deactivating Correlation Rules correlation score The possible values can be ALERT_COUNT each additional alert above MinAlertCount adds to the score SCENARIO_COUNT each distinct scenario starting with the se
87. expected based on the volume of data and system configuration Setting it very high will not affect performance since the call will return as soon as it is complete 600 Log UseDateLog WatchListScannerClass value Overrides the default Log UseDateLog property Identifies the Java class used to scan a watch list for a given name FALSE MantasWatchListScanner NameMatcherClass value Identifies the Java class used to match a name against a list of names FuzzyNameMatcher FuzzyMatcher SecondToPoll Identifies the number of seconds to wait between querying the WATCH_LIST table for new names that are added by the Watch List Management Utility FuzzyMatcher MaximumAddedNam es value Identifies the maximum number of names that can be added to the Watch List Service after it is initialized If additional names need to be added the service needs to be re initialized SSAParams System This corresponds to the Informatica Identity Resolution system to be used by Oracle Financial Services In practical terms this corresponds to the name of a subdirectory under the pr directory in the Informatica Identity Resolution installation For the purposes of Behavior Detection any name can be used but default is the standard Standard Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 101 Data Ingestion Directory Structure Chapter 4 Data Ingestion
88. fields in the account om app balances and position summary table The algorithm is a simple mathematical difference between two sums from two different queries plus the risk of straddled pairs Updates are made to assignment risk attributes 1020 w ph1020 update ap uncovered optio This workflow processes account option position pair ns from app data and updates the corresponding account position records Updates are made to the attributes relating to uncovered option contracts 1040 w ph1040 update abps option mktval This workflow processes account option position pair long from app data and updates the corresponding account balance records Updates are made to option market value long attributes 1050 w ph1050 aggregate ap to abps This workflow aggregates current day security positions by product category and account for update of the account balance record 1060 w ph1060 aggregate abps to hhbps This workflow aggregates account balance data into household balances 1070 w ph1070 aggregate ap to hhbps This workflow aggregates account positions data by household and updates the corresponding household balances record Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 317 Informatica Workflows Appendix D Informatica Workflow Details Table 120 Informatica Workflow Description Continued
89. follow these steps Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 13 About the Oracle Financial Services Job Protocol Chapter 2 Behavior Detection Jobs 1 Modify the install cfg file for algorithms to identify the directory where override datasets are stored The file resides in the following directory install dir behavior detection algorithms MTS mantas cfg install cfg The dataset override is specified with this property kdd custom dataset dir Note Specify the directory using a full directory path not a relative path If you do not or this property is not in the install cfg file the system disables the dataset overrides automatically 2 Create the dataset override file in the specified directory with the following naming convention dataset lt DATASET ID gt txt Note The contents of the file should start with the SQL definition in KDD DATASET SQL TX This SQL must contain all of the thresholds still represented for example eMin Indiv Trxn Am 14 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Performing Dispatcher Tasks Chapter 2 Behavior Detection Jobs Performing Dispatcher Tasks The dispatcher service runs on the server on which the application is installed Once the dispatcher starts it runs continuously unless a reason warrants shutting it down or it fails due to a problem This section describes the followi
90. for Alert Correlation Rule Migration Note In the install cfg file entries are in the form Property1 Property2 That is the value for Property1 is the value that processing assigns to Property2 As such if you change Property2 s value Property1 s value also changes Configuring the To configure the environment for alert correlation rule migration modify the Environment parameters that the sample INSTALL DIR database db tools mantas cfg install cfg shows Refer to Table 113 on page 290 The tables in the following sections describe the parameters specific to the Alert Correlation Rule Migration Utility Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 289 Alert Correlation Rule Migration Utility Chapter 8 Administrative Utilities Configuring General Table 113 describes general alert correlation rule migration parameters Alert Correlation Rule Migration Table 113 General Alert Correlation Rule Migration Parameters Parameter Description corrRuleMig CorrRule Location of the file containing the list of Alert Correlation FileNm Rule names to be migrated INSTALL DIR database db tools mantas cfg FileName cfg Note If the file name containing the list of Alert Correlation Rule Names is not provided the utility displays a warning message and extracts loads the default alert correlation rules specified in this file INSTALL DIR database db
91. for alerts such as the related Investment Advisor or Security involved in the alert Refer to section Update Alert Financial Data on page 176 for more information An alert scoring process ranks the alerts including each match within the alerts to indicate the degree of risk associated with the detected event or behavior Refer to section A rt Scoring on page 176 for more information An assignment process determines the user or set of users responsible for handling each alert The process assigns an alert to an owner based on customer specific assignment instructions Refer to section Assignment on page 177 for more information Optional An auto close algorithm closes alerts that are of a lower priority to the business The application automatically suppresses alerts according to customer specific auto close instructions Refer to section Awto Close on page 178 for more information Optional An auto supptession algorithm suppresses alerts that share specific scenario and focal entity attributes for a particular time frame The application automatically suppresses alerts according to customer specific auto suppression instructions Refer to section Automatic Alert Suppression on page 183 for more information The system generates highlights for alerts that appear in the alert list in the Alert amp Case Management subsystem and stores them to the database Refer to section Highlight Generation on page 184 for more info
92. for which placeholders are created in the PDF Template can be exported Caution Click Close button to close the report Closing the report using the browser close button will lock the report This would then prevent other users from editing the report for a short duration Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 315 Report Management Appendix C Report Management 316 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide APPENDIX D Informatica Workflow Details This appendix lists the Informatica workflow details used in the application and the use of that workflow Informatica Workflows Table 120 provides the list of workflows and a description for each Informatica workflow The workflows are listed by folders Table 120 Informatica Workflow Description Workflow Number Workflow Name Description BSM_Common Folder 0350 w ph0350 truncate rrst This workflow truncates the registered representative security table 0490 w ph0490 truncate hhsm This workflow truncate household summary month 0500 w ph0500 aggregate asm to hhsm This workflow aggregates household summary month out of the account summary month stage table 1005 w ph1005 truncate hhbps This workflow truncates household balance position table 1010 w ph1010 update abps asgmt risk fr This workflow calculates multiple
93. from mul tiple geographic regions for example Asia Pacific or Europe e Requires more processing power for a smaller batch window This window shrinks in multi regional deploy ment as batch does not start until data from all regions is available Alternative Process Multiple Sources Multiple processes that follow the same rules for self containment of data that single source processing follows For example transactions in one source are unable to reference an account in another source Processing of each batch of data occurs in sequential batches Behavior Detection does not allow overlap of batches Also you cannot combine processing of multiple sources with intra day batch processing Advantages e Spreads processing of multiple geographic regions across an entire day which contributes to fewer hard ware requirements e Makes alerts from a particular region available in a more timely manner Disadvantages e Makes processing more complicated to configure and monitor e Makes completion of processing of later batches in a timely manner more difficult when a delay of an ear lier batch occurs Live Market Data Mechanism for receiving market data from a live feed for example Reuters through an application such as TIBCO Advantages e Supports intra day processing e Enables an Oracle client to leverage its existing live feed infrastructure Disadvantages e Requires an expensive live feed infras
94. geographical areas in which the data in each batch does not overlap that is an execution in batch A does not occur against an order in batch B FALSE Processing does not separate data in each batch into a distinct time interval that is an event in batch A occurred at 10am and an event in batch B occurred at 9am and batch B arrived after batch A nj DT RegNMSExceptionCodes Identifies the Order Handling Codes that should be considered as Reg NMS executions ISO BAP BRD BOP SOE SHE FDT TreatLostEventsAsErrors value Identifies whether lost events found by the FDT refer to section Rejection During the Transformation Stage on page 110 for a discussion of lost events should be treated as errors TRUE or as lost events to be read in on the next run of FDT false TRUE FDT OpenOrderFileExpected value Identifies whether an OpenOrder file will be provided by the client during an end of day batch TRUE or whether it will not be provided FALSE TRUE FDT NonExecutionTradePurpose Codes value FDT DeriveTradeBlotter value Specifies a comma separated list of trade purpose codes For Trade Execution records that refer to an Order and have one of these codes the FDT will create a Trade record rather than an Execution record Specifies whether or not the FDT will create a TradeBlotter file CLNT BTBX FALSE FDT EnableMIFID value Identifies whether MiFid
95. in order to prevent conflicts and errors This section describes the recommended best practices for scenario migration for any existing Oracle Financial Services Behavior Detection system Caution Not following the recommended best practices while loading scenarios to the targeted system may cause one or more sequence ID conflicts to occur and your scenario will not be loaded Once a conflict occurs the metadata in the target environment needs to be corrected before the scenario can be successfully loaded To execute the recommended best practices you should have an intermediate level knowledge of the scenario metadata and be familiar with scenario patterns thresholds threshold sets and so on Basic SQL are required as well as access privileges to the MANTAS schema You must also be able to update records through SQLPLUS or a similar DB utility Scenario metadata is stored in many tables with each table using a unique sequence ID for each of its records If scenarios thresholds and scoring rules are modified in multiple environments using the same sequence ID range then conflicts may occur when you migrate scenarios to these environments To prevent conflict you must set different sequence ID ranges in each of the environments The recommended best practices contain two basic points e Make changes in only one environment e Separate the sequence ID ranges Prepare to implement the recommended best practices before installing the
96. information for the alert correlation rules to be extracted corr rule name in the someFileName cfg file and or lt corr_ rule name corr rule file name gt Note Providing both corr rule name and corr rule file name in the lt someFileName gt cfg file allows the user a flexibility to specify the actual filename that contains the metadata information for the respective alert correlation rule If you provide both an alert correlation rule name and an alert correlation rule file name on a line you must separate them with an equals sign It is recommended that the alert correlation rule file name be specified without an extension Navigate to the following directory cd INSTALL DIR gt database db tools mantas cfg Edit the install cfg file to include the path for the above created file against the tag corrRuleMig CorrRuleFileNm Navigate to the following directory cd INSTALL DIR gt database db tools bin Execute the sm extract sh script as follows sm extract sh correlation Note The utility performs the following validations upon extraction of an alert correlation rule The attribute value for type attribute in the XML tag Case should exist in the CASE TYPE CD column of the CASE schema table KDD CASE TYPE SUBTYPE The attribute value for subtype attribute in the XML tag Case should exist inthe CASE SUB TYPE CD column of the CASE schema table KDD CASE TYPE SUBTYPE The at
97. lectica 35 Table 14 KDD_BUS_DMN Table Attributes zucca cgi oet tia ies chain 37 Table T5 KDD ORG Table Attributes onesinddiepeputossditeo Dtegdce boton creme 41 Table 16 KDD_USER Table Attributes iius ecdeie t ctn de teer cia ci iit 46 Table 17 KDD USER ROLE Table Attributes eene 46 Table 18 System Account Passwotds snerist insisi piridi seisi 47 Table 19 Data Files by Gtoup iiie tte rrt pO er rh epa 58 Table 20 Preprocessing Output Directories sssssssseeeeeeeettes 61 Table 21 runMDS sh and runDP sh Output Directories sss 63 Table 22 runFDT sh Output Directoties seen 66 Table 23 runMDTish Output Directories 67 Table 24 Processing Batch Table Set up ceecsseesseseesessssessssessssescstsseensseenssesnsseenens 74 Table 25 Informatica Ingestion Parameters sese 75 Table 26 Data Ingestion Options and Alternatives TI Table 27 Directory Structure Description 81 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide xvii List of Tables Table 28 Table 29 Table 30 Table 31 Table 32 Table 33 Table 34 Table 35 Table 36 Table 37 Table 38 Table 39 Table 40 Table 41 Table 42 Table 43 Table 44 Table 45 Table 46 Table 47 Table 48 Table 49 Table 50 Table 51 Table 52 Table 53 Table 54 Table 55 Table 56 Table 57 Table 58 Table 59 Table 60 Table 61 Table 62 Table 63 Table 64
98. names ptior to correlation rules extraction or loading The Alert Correlation Rule Migration Utility creates and migrates the following metadata file lt CorrelationRuleName gt xml This file contains correlation rule metadata and additionally an audit trail of the correlation rule for core Oracle Financial Services Behavior Detection tables To avoid accidental loading of correlation rules into the incorrect environment the Alert Correlation Rule Migration Utility enables you to mame your source and target environments On extract you can specify the environment name to which you plan to load the correlation rule If you attempt to load it to a different environment the system displays a warning prompt The Alert Correlation Rule Migration Utility produces two log files Figure 65 on page 289 load log and extract log These files reside in the following location INSTALL DIR database db tools logs Using the Alert Correlation Rule Migration Utility Configuring the Alert Correlation Rules Migration Utility This section covers the following topics which describe configuring and executing the Alert Correlation Rules Migration Utility including extracting and loading metadata e Configuring the Alert Correlation Rules Migration Utility e Extracting Alert Correlation Rule e Loading Alert Correlation Rule The INSTALL DIR database db tools mantas cfg install cfg file contains common configuration information that Alert
99. outside the product range extract product range only N extract product range check N HHHH LOAD Specify the jdbc driver details for connecting to the target database load conn driver database driverName load conn url utils database urlName Target System ID load system id Ti5010S10 Specify the schema names for Load load schema mantas schema mantas owner load schema business schema business owner load schema market schema market owner load user miner utils miner user load miner password utils miner password Directory where scenario migration files reside for loading load dirname users mantast Solaris10 mantas58 b09 Ti5010810 Iron 13080 WAS database db tools data Specify whether threshold can be updated load threshold update Y Specify whether or not to verify the target environment on load verify target system N Continued on next page Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 201 About Administrative Utilities Chapter 7 Batch Processing Utilities Continued from previous page HHHHHHEHHHHHHEHH ALERT PURGE CONFIGURATION H HHHHHHHHHEHEHHAHEHEE Set the Alert Purge input variables here use the word null as the value of any parameters that are not to be used limit matches N purge Y batch size 5000 job null Scenario null enter dates with quotes in the following format DD MON YYYY HH24 MI S
100. particular user tasks and authorities Creating Scenario Manager Login Accounts To Create the Database Login Account As administrator the user setup process requires that you complete the following tasks 1 Create a database login account and password Refer to section To Create the Database Login Account on page 45 for more information 2 Setup an account and functional roles in the Scenario Manager Before performing any tasks in the Scenario Managet you must set up a user login account that establishes access and roles in the Scenario Manager Perform these setups by adding records to the database Refer to section To Set Up an Account and Functional Roles on page 46 for more information 3 Grant the database roles that the functional roles require You can grant the role of Data Miner or MNR to an Oracle Financial Services Scenario Manager user Refer to section To Grant a Database Rok on page 46 for more information Note Oracle Financial Services suggests having only a few generic users in the database to use the Scenario Manager as most organizations have an extremely small user population to execute these tools The system instantiates the Behavior Detection database as a set of Oracle database tables Therefore each user whom the Oracle client authorizes to use the Scenario Manager must have login access to the Oracle database As administrator you must set up an Oracle login account for each user and assign
101. provides sample configuration parameters fuzzy name fuzzy name fuzzy name fuzzy name fuzzy name fuzzy name fuzzy name fuzzy name fuzzy name fuzzy name fuzzy name fuzzy name fuzzy name fuzzy name fuzzy name fuzzy name fuzzy name fuzzy name fuzzy name fuzzy name fuzzy name fuzzy name fuzzy name fuzzy name match multi true file delimiter default prefix P vu vu vu U U U U tu oU Uo Ut tU UD Ut UJ Ut UJ max threads 1 max names per thread 1000 max names per process 250000 Stopword file mantas ingestion manager fuzzy match share stopwords b dat match threshold 80 initial match score 75 0 initial match p1 2 initial match p2 1 extra token match score 100 0 extra token min match 2 extra token pct decrease 50 first first match score 1 0 Stopword file mantas ingestion manager fuzzy match share stopwords p dat match threshold 70 initial match score 75 0 initial match p1 2 initial match p2 1 extra token match score 50 0 extra token min match 2 extra token pct decrease 50 first first match score 0 Figure 27 Sample install cfg Configuration Parameters Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 123 Fuzzy Name Matcher Utility Chapter 4 Data Ingestion Table 42 describes the utility s configuration parameters as they appear in the install cfg file Note that all scores have
102. raw data files from the Market data feed The logs subdirectory contains a log file for each component running on a host computer Each log file in the Logs subdirectory appears in a subdirectory with the date as its name in the format YYYYMMDD where Yyyy is the four digit year MM is the two digit month and DD is the two digit day The subdirectory s date is based on the processing date for data to which the log files pertain The IMC utility runIMc sh cleans up the logs subdirectory The IMC utility s configuration file defines the number of days that log files age before their removal Table 38 identifies log files for each component based on the file name s prefix Table 38 Log Files Output by Component Prefix Component XDP Preprocessor XDL Data loader MDS Market Data Server FDT File Data Transformer MDT Market Data Transformer TibW TibSpool TibW TibWatch IMC IMC Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 107 Startup and Shutdown Chapter 4 Data Ingestion Startup and Shutdown This section discusses Data Ingestion subsystem startup in both normal and recovery modes You can start and stop all components manually with their respective run and stop scripts with the exception of some components when configured to run in batch mode Given the complexity of Data Ingestion processing Oracle Financial Services recommends that a c
103. related data will be provided TRUE or not FALSE FALSE 96 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Ingestion Directory Structure Chapter 4 Data Ingestion Table 32 Datalngest xml File Configuration Parameters Continued me value file in the datamaps directory time a trade occurs during which any reported sales records cannot have the same price and quantity as the given trade to be considered as a market reference price 1 means that this condition will not apply 0 means the condition applies to reported sales with the same time 5 means the condition applies to reported sales within 5 seconds of the trade and so on This parameter is only used if FDT IgnoreFutureMarketRefs value FALSE Property Name Description Example FDT IgnoreFutureMarketRefs Identifies whether the FDT will use Reported FALSE value Market Sales records that occur later in time than a given trade when calculating the market reference price for that trade FALSE or not TRUE FDT MaxFutureMarketRefCompTi Specifies the number of seconds from the 1 The next four parameters are used to generate records in the TRADE TRXN CORRECTION table which record when a correction to a field of an execution trade or order occurs The fields to be checked for corrections should be specified in a comma separated list of business field names Business field names can be fou
104. result file and target name list file 124 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Fuzzy Name Matcher Utility Chapter 4 Data Ingestion Table 42 Fuzzy Name Matcher Utility Configuration Parameters Continued Parameter Description fuzzy name min intersection first letter count Specifies the number of words per name whose first letters match For example if parameter value 1 only the first letter of the first or last name would have to match to qualify If the value 2 the first letter of both the first and last name would have to match to qualify Warning By default the value is set to 2 Oracle Financial Services recommends using the default value You must not change the value to 1 or your system performance may slow down fuzzy name default prefix For entries that are not specified as business or personal name default to this configuration set fuzzy name max names per process This property variable determines whether or not the fuzzy matcher algorithm will be run as a single process or as multiple sequential processes If the total number of names between both the candidate name list and the target name list is less than the value of this property then a single process will be run If the number of names exceeds this property s value then multiple processes will be run based on how far the value is exceeded For example if the candidate nam
105. rules e Correlated Alerts by Business Entity Groups alerts created in the past month based on a common correlated business entity For example this rule would correlate all alerts with a business entity to alert correlation on customer CU122345 that were created in the past month e Potential Identity Theft Groups alerts created in the past seven days that ate generated on one or more specified scenarios where the alerts share a common correlated business entity Specified scenarios are those scenarios which identify behaviors that in isolation or when considered as a whole may be indicative of identity theft For example this rule would correlate all alerts generated on one or more of the specified scenarios with a business entity to alert correlation to CU122345 that were created in the past seven days The application installs these sample alert correlation rules in the INSTALL DIR database db tools data directory Account Approval Pre Trade Approval Tasks If the Oracle client has implemented the Account Approval AA Pre Trade Approval PTA application the following processes must be executed to successfully utilize both AA and PTA e Load Employee User Groups Captures the employees of the Oracle client along with their system logons It maps each employee to the Employee User Group which gives each employee access to the AA and PTA application to submit new account approval and pre trade approval requests respectively
106. s integration boundaries provide an Authorization header form field with principal or embedded principal to the Web Application server through a Web server plug in The tool also passes the same user IDs that the Oracle Financial Services directory server uses About Oracle Financial Services Operations As the Oracle Financial Services administrator you can coordinate the overall operations of the application Data Ingestion Behavior Detection and Post Processing Figure 4 Ir Ten Tne Poser Network Creation rsa Ca Match Scoring Alert Financial Alert Correlation Highlight Generation End Batch Figure 4 Oracle Financial Services Architecture Oracle Financial Services Processing In a production environment an Oracle client typically establishes a processing cycle to identify occurrences of behaviors of interest that is scenarios on a regular basis As Figure 4 illustrates each cycle of the Oracle Financial Services process begins with Data Ingestion Behavior Detection and Post Processing which prepares the detection results for presentation for the users Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 7 About Oracle Financial Services Operations Chapter 1 The Oracle Financial Services Behavior Detection Platform Start Batch Data Ingestion Behavior Detection Several factors determine specific scheduling of these processing cycles including availability o
107. sections describe these subsystems The following file is the message library for the Administration Tools application lt FIC_HOME gt AM admin_tools WEB INF classes conf mantas_cfg etc manta s admin tools message lib en dat All messages numbers that this log contains must be within the range of 50 000 89 999 The following file is the message library for the Database INSTALL DIR database db tools mantas cfg etc mantas database message lib en dat All messages numbers that this file contains must be within the range of 250 000 289 999 The following file is the message library for the Scenario Manager INSTALLED DIRECTORY behavior detection toolkit mantas cfg etc mantas toolkit message lib en dat All messages numbers that this section contains must be within the range of 130 000 169 999 The following file is the message library for the Services INSTALL DIR services server webapps mantas WEB INF classes conf mantas cfg etc mantas alert management message lib en dat All messages numbers that this section contains must be within the range of 210 000 249 999 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 301 Alert Management Case Management Appendix A Logging Alert Management Case Management The following logs contain the message library for both Alert and Case Management applications Web server Logs The following file is the message library for the W
108. set to Y creates Daily Partitions for the next week For example if run on a Thursday the DRM creates the five 5 partitions for the next week beginning with Monday Caution For weekly partitioned tables do not set the value to Y KDD DR JOB Table The KDD DR JOB table stores the start and end date and time and the status of each process that the Data Retention Manager calls Table 98 provides the table s contents Table 98 BUSINESS KDD DR JOB Table Contents Column Name Description JOB ID Unique sequence ID START DT Start date of the process END DT End date of the process STATUS CD Status of the process e RUN Running FIN Finished successfully e e ERR An error occurred e WRN Finished with a warning 250 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Database Statistics Management Chapter 7 Batch Processing Utilities KDD DR RUN Table The KDD DR RUN table stores the start and end date and time and status of individual process runs that are associated with a table Table 99 describes the table s contents Table 99 BUSINESS KDD DR RUN Table Contents Column Name Description JOB ID Unique sequence ID PROC ID Process ID START DT Start date of the process END DT End date of the process RSULT CD Result of the process e FIN Finished successfully e ERR An error occurred e WRN Finished with a warning
109. specifies e Whether to create Employee Trading Restriction records for all employees in the organization and all the related child organizations defined in the Organiza tion Relationship file TRUE or e Whether to create records only for employees in the specified organization False FALSE XDP Data Type gt lt Property gt Overrides the given property for the given Preprocessor instance XDL Specifies properties used to configure the Data Loader XDL component XDL Default Refresh Is valid for data types that have a load operation of Overwrite as defined in the DIS This parameter specifies replacement of the entire table TRUE or provision of deltas FALSE TRUE XDL Default DataFileExts Specifies the possible file extensions for an input file XDP FDT MDT XDL Default CommitSize XDL Default ErrorLimit Specifies the number of records to update or insert before committing not used when Direct TRUE Specifies the number of rejected records to allow before exiting with an error If left blank the default processing sets no limit 500 XDL Default DbErrorCodes Specifies a comma separated list of database vendor specific error codes that indicate data level errors for example data type and referential integrity This results in rejection of records with a warning instead of a fatal failure 1 140 014 O11 407 140 000 000 000 000
110. such as WebSphere WebLogic Linux Oracle and Informatica Refer to the Oracle Financial Services Behavior Detection Platform Stage 1 Installation Guide for more information about these tools Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 1 About the Oracle Financial Services Architecture Chapter 1 The Oracle Financial Services Behavior Detection Platform About the Oracle Financial Services Architecture An architecture is a blueprint of all the parts that together define the system its structure interfaces and communication mechanisms A set of functional views can describe an architecture The following views illustrate the implementation details of the application s architecture e Component View Illustrates system components and their dependencies e Deployment View Illustrates the deployment of components to processing nodes e Security View Emphasizes the security options between processing nodes through a specialized deployment view The following sections describe these views Component View Figure 1 illustrates the concept that a series of tiers and subsystems compose the application s architecture as a MULA G A Alert amp Case Management Behavior Detection Data Ingestion Figure 1 Oracle Financial Services Architecture Overview 2 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide About t
111. table may look similar to the example in Table 84 Table 84 Sample KDD PRCSNG BATCH Table with Single Batch PRCSNG BATCH NM PRCSNG BATCH DSPLY NM PRCSNG ORDER EOD BATCH NM DLY Daily Batch 1 DLY Single Site Intra day In this intra day batch example the system is servicing a single time zone but runs an Processing additional batch during the day to identify behaviors related to overnight trading as Table 85 desctibes Table 85 Sample KDD PRCSNG BATCH Table with Intra day Processing PRCSNG BATCH NM PRCSNG BATCH DSPLY NM PRCSNG ORDER EOD BATCH NM MAIN Main Evening Batch 2 MAIN MORN Morning Batch 1 MAIN In this configuration run the Calendar Manager Utility only during the MORN batch Refer to Calendar Manager Utility on page 236 for more information You can run the Data Retention Manager in either the MORN or MAIN batch If you run it in the MAIN batch define at least one buffer partition so that the MORN batch does not fail due to inadequate partitions Refer to Data Retention Manager on page 241 for more information 226 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Batch Control Utility Chapter 7 Batch Processing Utilities Multiple Countries A single deployment supports detection against data from New York London and Hong Kong In this case three batches are all end of day batches as Table 86 describes Table 86
112. text within the message itself Under normal circumstances the message IDs are not subject to change between releases but the text of the message can change If a message ID does change you can refer to the appropriate readme txt file for information about updated IDs Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 309 Logging Configuration File Appendix A Logging 310 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide APPENDIX B Oracle Financial Services Software Updates This appendix describes the application of Behavior Detection Platform software updates in Oracle Financial Services Behavior Detection Platform Oracle Financial Services Software Updates Hotfix Hotfix Effect on Customization Behavior Detection Platform Software Updates Hotfix A hotfix is a package that includes one or more files that are used to address a defect or a change request in Oracle Financial Services Behavior Detection Typically hotfixes are small patches designed to address specific issues reported by the clients Hotfixes can affect the following areas in Behavior Detection The User Interface UD Scenarios patterns and datasets Post Processing jobs Performance Informatica Ingestion Each hotfix includes a readme txt file which describes the step by step process to install the hotfix Hotfixes are delivered to clients in the following ways E mail Se
113. that were not processed due to error data errors Directory Archive Specifies the parent directory for directories that contain backup copies of intermediate files that Java ingestion components create data backup Directory Config Specifies the directory containing configuration files for Java ingestion server config Directory FuzzyMatcher Specified the directory containing files related to fuzzy matcher fuzzy match data files that the Oracle client submits The default value of dat is in accordance with the DIS Directory DataMap Specifies the directory that contains XML config datamaps data map files FileExtension Specifies properties used to define extensions for various types of files FileExtension Log Specifies the file name extension for log files 1og FileExtension Checkpoint Specifies the file name extension for Cp checkpoint files Many of the Java ingestion components create checkpoint files as an aid to recovery when restarted after exiting prematurely FileExtension Temporary Specifies the file name extension for tmp temporary files that Java ingestion components create FileExtension Error Specifies the file name extension for error err files that Java ingestion components create FileExtension Data Specifies the file name extension for input dat Separator value Specifies the delimiter that separates fields in data file records
114. the Behavior Detection assigned security identifier SecurityISIN KeyColumn Specifies the column name to use when SCRTY INTRL ID looking up security records by Behavior Detection assigned security identifier SecurityISIN ValueColumn Specifies the column to retrieve when looking SCRTY ISIN ID up a security using the Behavior Detection assigned security identifier SecurityISIN Category Specifies the category of data in which the BUSINESS security table resides The category is a key for mapping to the database schema in which the security table resides MDS Specifies properties used to configure the MDS component MDS NumberOfThreads value Specifies the number of worker threads that 4 the MDS uses when processing data MDS BufferSize value Allows an override to the 1024 BufferSize value property for MDS MDS Adapter value Specifies the type of market data feed being Reuters provided by the client Reuters indicates processing of a live market feed using TIBCO software File indicates that the Oracle client provides market data as flat files in accordance with the DIS 92 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Ingestion Directory Structure Chapter 4 Data Ingestion Table 32 Datalngest xml File Configuration Parameters Continued Property Name Description Example MDS Properties for Use with Liv
115. the KDD_MNR user role to this account Note Behavior Detection does not support external logins for example OPS accounts in an Oracle environment Users must provide an explicit password when logging on The assumption is that the client s system administrator has training and experience in performing such setups and therefore does not require instructions here on how to perform this task However for information about setting up Oracle accounts refer to the appropriate Oracle documentation Note The Solaris and Oracle login user IDs do not have to be identical However the Oracle Financial Services Scenario Manager and Oracle login user IDs MUST be identical Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 45 About Scenario Manager Login Accounts Chapter 3 Security Configuration To Set Up an Account To create a Scenario Manager account and functional role follow these steps and Functional Roles 1 Access the KDD_USER table Table 16 defines the attributes for the KDD_USER table Table 16 KDD_USER Table Attributes Column Name Description USER_ID User s database login ID USER_NM User s name USER_ROLE CD User s default database role ACTV_FL Active user indication Y or N WRKLD CD Not used by the Scenario Manager 2 Enter the following information into the table using an SQL script a User database login ID in the USE
116. to clear all job logs 4 Restart the dispatcher Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 23 Recovering Jobs from a System Crash Chapter 2 Behavior Detection Jobs Recovering Jobs from a System Crash If the system crashes all active jobs status_cd RUN fail You can recover the jobs by running the script recover_mantas sh This script changes the status_cd to RES so that these jobs can restart and finish running The recover_mantas sh script has an optional parameter the date on which the system ran the start_mantas sh script This parameter has a DD MON YYYY format The default value is the current date Running the recover_mantas sh script with this parameter ensures the script recovers only the jobs started that day The dispatcher must be running to pick up the restarted jobs This results in either a successful completion status_cd FIN or failure status_cd ERR You can restart jobs that ended in failure by running the restart_mantas sh script The restart_mantas sh template group id script changes the status cd from ERR to RES for any jobs passed in the template group that have a status cd of ERR for the dispatcher to pickup 24 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide CHAPTER 3 Secunty Configuration This chapter provides instructions for setting up and configuring the Security Management System SMS to support u
117. vs TOOLS re TOOLS re TOOLS re LS re MP NETACENCU NJ MP NETACENCU NT MP NETACENCU Ni Fresh temp Fresh Fresh temp Fresh temp P NETACENCU NT ACCTPSWRD tabl Fresh temp NETACENCU NT ACCTTAXID tabl Fresh temp D ACCTPHONE tabl tabl le sh T ACCTADDR VW tabl le sh VW tabl le sh VW le sh VW le sh VW le sh P NETACENCU NT CUACADDR VW P NETACENCU NT P NETACENCU NT P NETACENCU NT P NETACENCU NT r ACCTADDR T ACCTEMAIL T ACCTPHONE TMP NETACENCU NT ACCTPSWRD P NETACENCU NT ACCTTAXID r CUACADDR 256 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Refreshing Temporary Tables Chapter 7 Batch Processing Utilities M T T T T T T T DB TOOLS re MP NETACENCU NT D vs TOOLS re Le DB TOOLS re MP NETACENCU NT D pam MP NETACENCU NT D amm MP NETACENCU Nl D MP NETACENCU NT v TOOLS re v TOOLS re v TOOLS re DB TOOLS re is 114698120 NETACENCU NT Fresh temp Fresh temp Fresh temp Fresh temp JRNL VW Fresh temp Fresh temp table sh CUACEMATL VW table sh CUACPHONE VW table sh CUACTAXID VW table sh table sh D WIREACBENE VW table sh D WIREACORIG VW i Fresh temp table sh MP NE
118. xml Analysis configuration specific for data displayed in support of Anti Money Laundering scenarios analysis iaml xml Analysis configuration specific for data required by Institu tional Anti Money Laundering scenarios and Ingestion Manager operations to support them analysis iaml ui xml Analysis configuration specific for data displayed in support of Institutional Anti Money Laundering scenarios analysis bc xml Analysis configuration specific for data required by Broker Compliance scenarios and Ingestion Manager operations to support them analysis bc ui xml Analysis configuration specific for data displayed in support of Broker Compliance scenarios You can also create your own files using the provided files as a template Place files that you create in the mantas cfg directory that the DTD can locate If you place your files in a different directory you must modify the DTD reference in the XML files to qualify the path to the DTD To Run the Data Go to the install dir database db tools bin directory and execute the Analysis Tool following command run data analysis tool sh bg i input file xml o outputfile Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 271 Data Analysis Tool Chapter 8 Administrative Utilities Table 103 describes the command line arguments that the Data Analysis Tool uses Table 103 Command Line Arguments Argument Explana
119. 000 000 000 The following properties apply only to the Oracle adapter XDL Default MaxBindSize Specifies the maximum number of bytes integer to use in the bind array for loading data into the database 4194304 XDL Default Direct Specifies whether to use direct path loading TRUE or conventional path loading FALSE FALSE XDL Default Parallel Specifies whether a direct path load will be done in parallel TRUE This will be the case when multiple loaders for the same data type are run in parallel such as with multiple ingestion instances FALSE Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 99 Data Ingestion Directory Structure Chapter 4 Data Ingestion Table 32 Datalngest xml File Configuration Parameters Continued Property Name Description Example XDL Default Unrecoverable Specifies whether a direct path load does not FALSE use redo logs TRUE or uses redo logs FALSE XDL Default Partitioned Specifies whether a direct path load uses the FALSE current date partition TRUE or any partition FALSE XDL Default SkipIndexes Specifies whether a direct path load skips FALSE index maintenance TRUE or maintains indexes FALSE If set to TRUE rebuilding of indexes must occur after running the Data Loader XDL Default SkipIndexErrorCo Specifies a database vendor specific error 26025 de code that occu
120. 1 Current Day P20050708 Previous Day and End of week 1 P20050701 End of previous week 2 P20050630 End of previous Month 1 P20050624 End of previous week 3 P20050617 End of previous week 4 P20050531 End of previous Month 2 To execute Data Retention Manager use the following procedure Be sure to run the utility when usets are not working on the system To avoid conflicts Oracle Financial Services recommends that you use this utility as part of the end of day activities The Data Retention Manager should be executed nightly for Daily partitioned and Mixed date partitioned table after the calendar has been set for the next business day For weekly and monthly partitioned table the Data Retention Manager should be executed prior to the end of the current processing period Oracle Financial Services Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 245 Data Retention Manager Chapter 7 Batch Processing Utilities recommends running the Data Retention Manager on Thursday or Friday for weekly partitioned tables and on or about the 23rd of each month for monthly partitioned tables Note Be sure to set the system date with the Calendar Manager Utility prior to running the Data Retention Manager Refer to Calendar Manager Utility on page 236 for more information To Run the Data To run Data Retention Manager manually follow these steps Retention Manager
121. 114 Table 115 Table 116 Table 117 Table 118 Table 119 Table 120 Environment 1 Development eee sse 284 Environment 2 Iest WAT s dede ettet ee ge eene 285 B vironment 3 PROD i eee ret a rtt teorie rodeo eqs 285 General Alert Correlation Rule Migration Parametets 290 Alert Correlation Rule Extraction Parameters 290 Alert Correlation Rule Load Parameters sse 291 lernt d M 300 Logging Location Property Values eee 307 Configurable Parameters for Common Logging ess 308 Metadata Creation for Report Framewotk sse 314 Informatica Workflow Description seen 317 Xx Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide About this Guide This guide explains the concept behind the Oracle Financial Services Behavior Detection Platform and provides comprehensive instructions for proper system administration as well as daily operations and maintenance This section focuses on the following topics Who Should Use this Guide Scope of this Guide How this Guide is Organized Where to Find More Information Conventions Used in this Guide Who Should Use this Guide The Oracle Financial Services Behavior Detection Platform Administration Guide is designed for use by the Oracle Financial Services Installers and System Administrators Their roles and responsibilities as t
122. 1502_Risk_Review_Trusted_Pair Risk review only happens if managing tp from uiis set to Y in the installMantas properties sample properties file Sequence Number Folder Workflow Number Workflow Name Predecessor 13 ORION Common 2172 w ph2172 update jurisdiction in ag 13 ORION Common 2200 w ph2200 build trxn tables from fotps ORION Common 3051 ORION Common 3210 14 ORION Common 2220 w ph2220 update fot reversals ORION Common 2200 20 ORION Common 3501 w ph3501 Exp and Risk Review TP ORION Common 2200 21 ORION Common 3502 w ph3502 Flag Trusted Trxn ORION Common 3501 142 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide AML Brokerage Workflows Chapter 5 Informatica Workflows Informatica Summary Workflows AML Brokerage Informatica Summary workflows maintain monthly aggregations of customer activity Figure 32 illustrates Informatica Summary workflows for AML Brokerage External 2110 3210 0371 Predecessor Watch List Watch List Watch List ied Sequence 16 0210 0220 0230 Sequence 17 Ld p ue Sequence 18 Sequence 19 Sequence 20 0280 N P sequence dependent processing can occur simultaneously across all sequences Sequence 21 0760 Workflows with predecessors must wait for predecessor results before workflow processing can begin Figure 32 Summary Workflows AML Brokerage The application completes these
123. 1a and 701b this parameter controls how many days of transactions to look across Verify whether the new data contains rever sals of prior transactions Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 75 Informatica Ingestion Parameters Chapter 4 Data Ingestion Table 25 Informatica Ingestion Parameters Continued File Variable Description prod_orion parm MIN GEO RISK Defines what is considered High Risk For Continued the Account Profile attributes related to High Risk Geography for example Incoming High Risk Wire Count Processing compares this parameter using a strict greater than operation S SBASE COUNTRY Sets the country for the derived institution when deriving institution records from transactions if no address information is provided on the transaction SSPROCESS PASS THRU Identifies whether data ingestion determines if a transaction is pass through or if the client is performing this determination e Y indicates that the Behavior Detec tion mapping runs e Nindicates that the firm calculates this field and provides the value during ingestion SPROCESS SECONDARY NAMES Tells data ingestion whether it should populate the secondary originator and secondary beneficiary name fields in the Front Office transaction table or if it should leave it blank for the firm to populate e Y indicates that the Behavior Detec tion mapping runs
124. 2 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Batch Export Utility Chapter 7 Batch Processing Utilities Script Details Example Usage Attachments XML Metadata Table 90 Alert Metadata The run pdf archival utility sh script default calls the Investigation Export Web service which runs under the Alert Management UI to export alert and case information This sctipt exports all alerts and cases that were closed after the last execution of the utility without any parametets set run pdf archival utility sh ALERT_ID lt comma separate alert ids CASE ID comma separate case ids gt The ALERT ID and CASE ID are optional If none of these are passed to the script run pdf archival utility sh then the alerts and cases that were closed after the last execution of the script are archived as in the following example run pdf archival utility sh When either an ALERT ID or CASE ID or both are passed then the corresponding alerts and cases are archived as in the following examples 1 run pdf archival utility sh ALERT ID 1234 3435 4354 3454 2 run pdf archival utility sh CASE ID 8999 8984 9800 8498 3 run pdf archival utility sh ALERT ID 1234 3435 4354 3454 CASE ID 8999 8984 9800 8498 Note Running the script with ALERT ID or CASE ID or both does not update the last execution date The ALERT ID and CASE ID should be numbers only Use KDD REVIEW REVIEW ID values as
125. 20000000 29999999 KDD TSHLD SET TSHLD SET ID SEQ 20000000 20000000 29999999 Table 112 Environment 3 PROD TABLE NM SEQUENCE NAME CURRENT VALUE MIN VALUE MAX VALUE KDD ATTR ATTR ID SEQUENCE 30000000 30000000 39999999 KDD AUGMENTATION AGMNT INSTN ID SEQ 30000000 30000000 39999999 KDD DATASET DATASET ID SEQUENCE 30000000 30000000 39999999 KDD JOB JOB ID SEQ 30000000 30000000 39999999 KDD LINK LINK ID SEQ 30000000 30000000 39999999 KDD LINK ANLYS NTWRK DEF NTWRK DEFN ID SEQ 30000000 30000000 39999999 N KDD LINK ANLYS TYPE CD TYPE ID SEQ 30000000 30000000 39999999 KDD LINK SUMMARY LINK SUMMARY ID SEQ 30000000 30000000 39999999 KDD LINK TYPE SUMMARY LINK TYPE SUMMARY ID 30000000 30000000 39999999 SE KDD NODE NODE ID SEQ 30000000 30000000 39999999 KDD NTWRK NTWRK ID SEQ 30000000 30000000 39999999 KDD PARAM SET PARAM SET ID SEQ 30000000 30000000 39999999 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 285 Scenario Migration Utility Chapter 8 Administrative Utilities Table 112 Environment 3 PROD Continued TABLE NM SEQUENCE NAME CURRENT VALUE MIN VALUE MAX VALUE KDD PTTRN PTTRN ID SEQ 30000000 30000000 39999999 KDD RULE RULE ID SEQ 30000000 30000000 39999999 KDD SCNRO SCNRO ID SEQ 30000000 30000000 39999999 KDD SCORE SCORE ID SEQ 30000000 30000000 39999999 KDD SCORE HIST SCORE HIST SEQ ID SEQ 30000000 30000000 39999999 KDD TSHLD T
126. 2005 set cd pd dates 1 ORION Common 2050 w ph2050 update bot reversals 1 ORION Common 2070 w ph2070 truncate cas 1 ORION Common 2140 w ph2140 pass thru process 1 ORION Common 2180 w ph2180 instn identification 2 ORION Common 2190 w ph2190 fotps instn processing ORION Common 2180 2 ORION Common 2191 w ph2191 anticipatory profile instn processing ORION Common 2180 3 ORION Common 2130 w ph2130 update fot unrelated party code 3 MLM Brokerage C 2192 w ph2192 update inst instn seq id ORION Common 2180 ommon Informatica Watch List Workflows Fraud Detection Informatica Watch List workflows facilitate the application of customer supplied measures of risk to corresponding entities transactions and instructions 160 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Fraud Detection Informatica Workflows Chapter 5 Informatica Workflows Figure 46 illustrates Informatica Watch List workflows for Fraud Detection Note Processing of workflows is not sequence dependent processing can occur simultaneously across all sequences Workflows with predecessors must wait for predecessor results before workflow processing can begin Figure 46 Watch List Workflows Fraud Detection The application completes these workflows in 10 sequences processing of some workflows requires data from Miscellaneous predecessors These workflows may proceed simultaneously in all sequences as soon as
127. 248 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Retention Manager Chapter 7 Batch Processing Utilities Maintaining Indexes Utility Work Tables KDD DR MAINT OPRTN Table To Add a Monthly Database Partition To add a monthly partition run the utility s shell script as follows Refer to Table 95 for parameters run_drm_utility sh AM M BUSINESS lt object gt N where AM is the drm_operation parameter that implies adding a monthly partition To Drop a Monthly Database Partition To drop a monthly partition run the utility s shell script as follows Refer to Table 95 for parameters run_drm_utility sh DM M BUSINESS lt object gt N where DM is the drm operation parameter that implies dropping a partition As part of processing the Data Retention Manager automatically rebuilds the database index and index partitions that become unusable You do not need to maintain the indexes separately The utility enables you to rebuild global indexes by executing the following command run drm utility sh RI M BUSINESS object N where RI is the drm operation parameter that implies rebuilding indexes The Data Retention Manager uses three work tables during database partitioning which the following sections describe e KDD DR MAINT OPRTN Table e KDD DR JOB Table e KDD DR RUN Table The KDD DR MAINT OPRTN table contains the processing information that manages Data Retention
128. 273 Data Analysis Tool 261 262 265 about 262 analysis constraints 264 analysis XML files 263 265 configuring 262 join counts analysis 267 268 269 logs 272 Production Data Model 262 queries 270 report 272 schema 262 troubleshooting 273 using 271 Data Ingestion 8 51 about 51 alternatives 77 analyzing indexes 69 archiving 112 backup server configuration 108 cleanup activities 112 client inbox directory 57 60 configuration files 62 67 86 87 88 control data 129 data loaders 61 data rejection 109 110 112 data tables 69 directory structure 57 80 distributed processing configuration 109 events 111 Firm Data Transformer FDT 64 FSDM 69 Fuzzy Name Matcher Utility 51 120 Informatica parameters 75 Informatica workflows 133 135 146 153 158 166 Ingestion Manager 51 intra day processing 73 job scheduling 51 lost events 111 Market Data Server 63 Market Data Transformer MDT 64 optional miscellaneous workflows 136 146 153 158 166 out of sequence 111 Post Watch List Miscellaneous workflows 140 163 169 preprocessing 57 60 61 109 110 pre processing TCS data 63 Pre Watch List Miscellaneous workflows 137 pre watch list miscellaneous workflows 159 process flow 52 rebuilding indexes 68 recovery 108 scripts 57 63 64 67 69 82 85 133 startup and shutdown 108 subdirectories 80 82 85 103 104 105 106 107 summary workflows 164 Trading Compliance da
129. 3 Security Configuration Organization Jurisdiction Business Domain Scenario Class Case Type Subtype Correlation Rule Additional Parameters The Choose User drop down list filters its values based on the value selected in the Choose User Type selection drop down list It shows only users if the User Type is User and it shows only organizations if the User Type is Organization After selecting the desired user in Choose User drop down list the Administrator can map the following parameters to the selected user e Organization e Jurisdiction e Business Domain e Scenario Class e Case Type Subtype e Correlation Rule A User or Organization s access to other Organization depends on the selection s made for this organization parameter For Example if a user is mapped Orel and Org2 it implies that user can access alert case which belongs to these two organizations provided other security attributes are also matching Mapping of one or more jurisdictions to a user or organization gives the privilege of accessing cases alerts account approval requests or pre trade approval requests that belong to the mapped jurisdiction Mapping of one or more business domains to a user or organization gives privilege of accessing cases alerts account approval requests or pre trade approval requests that belong to the mapped business domains Mapping of one or more Scenario Classes to a user or organization gives the privil
130. 4 hours 6 Logs a message in the INSTALL DIR database db tools logs batch control 1log file stating that the batch process has begun At times you may need to know the name of a currently running batch or verify that a batch is active For example during intra day detection processing many batches may be running simultaneously and you need to identify one or more by name To identify a running batch process use the following procedure Caution If you set the batch control logging to display at the console be aware that log messages are mixed with the output of the shell script the output can be difficult to read To obtain a batch name follow the steps 1 Access the directory where the shell script resides cd INSTALL DIR database db tools bin 2 Start the batch shell script get mantas batch sh The name of the currently running batch is written to standard output Refer to Configuring Console Output on page 210 for more information After establishing the required Java environment and initiating various Java processing activities the Batch Control Utility does the following 1 The utility retrieves the name of the currently running batch from the KDD PRCSNG BATCH CONTROL table Refer to Table 87 on page 228 2 The utility returns the batch name to standard output Batch Export Utility The Batch Export utility allows Oracle Financial Services clients to export batches of alerts and cases for archi
131. 508 where 508 is the job template that Oracle Financial Services provides to run the Alert Correlation algotithm 186 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Alert Correlation Chapter 6 Post Processing Tasks Understanding Alert Correlation Configuration Business Entity Paths As mentioned above Alert Correlation performs two major tasks correlating alerts to business entities and correlating alerts to alerts The second step is optional and is governed by the correlate_alerts_to_alerts job parameter delivered with the template job associated to group 508 If this parameter s value is set to rze then this step will be performed and if this value is set to Jasse then it will not be performed The only exception to this is if the correlate alerts fo alerts feature is not licensed A license for this feature can be obtained from your engagement representative and details on enabling the license file can be found in the Oracle Financial Services Behavior Detection Platform Installation Guide Release 6 1 3 The other job parameter associated with Alert Correlation is corre ation actions This parameter has a value of a comma delimited list that defines what optional actions to take against a correlation that is found by the correlate alerts fo alerts task The currently supported actions are prioritize which will assign a score to the correlation and promote fo case which will prom
132. 6 1 3 Administration Guide Fraud Detection Informatica Workflows Chapter 5 Informatica Workflows These workflows may proceed simultaneously in all sequences as soon as processing of any predecessors has completed Table 65 describes the Summary workflows in Figure 48 Table 65 Summary Workflows Fraud Detection Sequence Number Folder Workflow Number Workflow Name Predecessor 16 ORION Production 0150 w p0150 truncate asms ORION Common 2005 16 9001 AccountDailyProfileTrade 16 9002 AccountDailyProfileTransaction 17 ORION Production 0190 w p0190 aggregate fotps to asms ORION Common 3210 ORION Production 0150 17 9005 AccountProfile 9001 9002 18 ORION Production 0310 w p0310 update asm for daily activity 9005 ORION Production 0190 19 ORION Production 0630 w p0630 build ATxSD from ASMS ORION Production 0310 19 ORION Production 0750 w ph0750 truncate APTxSM 19 MLM Brokerage P 0280 w p0280 create clog activity records ORION Production 0310 roduction 20 ORION Production 0760 w ph0760 ASM to APTxSM ORION Production 0750 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 165 Insurance Workflows Chapter 5 Informatica Workflows Insurance Workflows The following sections describe the Informatica workflows that are required for deriving and aggregating data for the Insurance Solution e Informatica Miscellan
133. 6 Post Processing Tasks sessessssssee 171 About Post Processttigu i cades atte ceste pete e tae d leprosi AT R 171 Order of Running Post Processing Administrative Tasks essent tette 173 Match SCOEItig sce ict opin teet terere OI EEI E ENON GI R tes 173 Running the Match Scoring Job itti eret esatto es tonne oet oie a hene i E aE dE deos 173 To Run the Math Scnne 0b5 tU NEE TREE IEEE nAn 173 Alert Creation iu detiene ee o eerte erri r ie rois e pe eo E ETET 174 Running the Alert Creation OD is sass odo oret eee e i eterni dto iei eo le dde 174 To Rum Mutt math Alert CTM0F aet nete tegi evt myrto n inei e o nemine tiae 174 To Runs mile Match Alert Creator s e edat tee OP REOS ETE eR Pe eode 174 Understanding Advanced Alert Creator Configuration ssssssseeee eene 174 Advanced RULES cia uaunttente etenim edt a tiri de OR n e fof roris 174 Update Alert Financial Datascscccccstsssssssisstsassissastssccssesetesscssssescvestasonanstvensovadiasenssusssiosusessdesevsevanstvanspnausiveestessdecusevedeeetoes 176 PoXdnvrue sit 176 Running the Alert Scotinig OW 5t reto Qetan Prem ons une resa ers vos ceste n date eroe de o aane cR 176 Serie 177 Running the Assignment Job encre one tei ere ESEI E ISEE bue Ii iS 177 fuil llc 178 D
134. 60 7 ORION Common 3150 w ph3150 load staging and validate watch list ORION_Common 3040 ORION Common 3070 ORION Common 3090 ORION_Common 3 091 ORION_Common 3100 MLM_Brokerage_Common 3120 MLM_Banking_Common 3130 8 ORION Common 3170 w ph3170 update staging list risk ORION Common 3150 9 ORION Common 3190 w ph3190 apply risk to nonacct entities ORION Common 3005 ORION Common 3170 ORION Common 3200 w ph3200 apply membership to entities ORION_Common 3170 11 ORION Common 3191 w ph3191 apply risk to acct entities MLM Brokerage Common 3171 ORION Comm 0on 3190 12 ORION Common 3210 w ph3210 update fotps activity risk ORION Common 3191 12 ORION Common 3240 w ph3240 update INS activity risk ORION Common 3191 168 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Insurance Workflows Chapter 5 Informatica Workflows Informatica Miscellaneous Post Watch List Workflows Insurance Post Watch List Miscellaneous workflows perform processing in support of workflows in multiple functional areas Figure 51 illustrates Post Watch List Miscellaneous Informatica workflows for Insurance External Predecessor 2200 Sequence 20 3501 Note Processing can occur simultaneously within each sequence Workflows with predecessors Sequence 21 must wait for predecessor 3502 results before workflow processing can b
135. 91 ORION_Common 3100 MLM_Brokerage_Common 3120 MLM_Banking_Common 3130 8 ORION Common 3170 w ph3170 update staging list risk ORION Common 3150 9 ORION Common 3190 w ph3190 apply risk to nonacct entities ORION Common 3005 ORION Common 3170 ORION Common 3200 w ph3200 apply membership to entities ORION Common 3170 11 ORION Common 3191 w ph3191 apply risk to acct entities MLM Brokerage Common 3171 ORION Comm 0on 3190 12 ORION Common 3210 w ph3210 update fotps activity risk ORION Common 3191 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 149 AML Banking Informatica Workflows Chapter 5 Informatica Workflows Informatica Miscellaneous Post Watch List Workflows AML Banking Informatica Miscellaneous Post Watch List workflows facilitate the application of customer supplied measures of risk to corresponding entities transactions and instructions Figure 37 illustrates Informatica Miscellaneous Post Watch List workflows for AML Banking Sicca RE 3130 sle Watch List Watch List Sequence 13 Sequence 14 Sequence 20 Note Processing can occur simultaneously within each sequence Sequence 21 Workflows with predecessors must wait for predecessor results before workflow processing can begin Figure 37 Post Watch List Workflows AML Banking The application completes these workflows in four sequences Pr
136. A drm_owner schema mantas owner drm object name A drm weekly proc fl Y Directory required to put the temporary log from Database Server This example shows default values that the system uses only when calling the utility with no command line parameters 244 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Retention Manager Chapter 7 Batch Processing Utilities Executing the Data Retention Manager Table 95 describes these parameters Table 95 Data Retention Manager Processing Parameters Parameter Description drm_operation Operation type P Partition AM Add Monthly Partition DM Drop Monthly Partition RI Rebuild Indexes RV Recompile Views T Truncate Current Partition drm partition type Partition type D Daily W Weekly M Monthly X Mixed Date A All Partitions Daily Weekly Monthly drm owner Owner of the object database schema owner drm object name Object name If performing an operation on all objects the object name is A drm weekly proc fl Flag that determines whether partitioning occurs weekly Y and N Note The system processes Daily partitioned tables drm partition type D and Mixed date partitioned tables drm partition type X simultaneously Therefore you need only specify D or X to process these tables An example for the Mixed date partition for the present date 20050711 is P2005071
137. AMCMLookupFiles 2 All date values should be provided in MM DD yyyy format in the Excel wotksheet 3 Whenever a record is deleted from the Excel the complete row should be deleted In other words no blank active record should exist in the Excel 4 After selecting the Excel template preview it before uploading The Excel Upload screen can be accessed by logging in as Admin user The Excel template to be used is KDD_JURISDICTION xls Creating Business Domain in the Database Business domains ate used for data access controls similar to jurisdiction but have a different objective The business domain can be used to identify records of different business types for example Private Client vs Retail customer or to provide more granular restrictions to data such as employee data The list of business domains in the system resides in the KDD_BUS_DMN table Behavior Detection tags each data record provided through the Ingestion Manager to one or more business domains Behavior Detection also associates users with one or more business domains in a similar fashion If a user has access to any of the business domains that are on a business record the user can view that record The business domain field for users and data records is a multi value field For example you define two business domains e a Private Client eb Retail Banking A record for an account that is considered both has BUS_DMN_SET ab If a user can view business domain
138. Account Address example given above the value would be Accountldentifier AddressPurpose If this parameter is not present then the primary key located in the given data type s data map file for example datamaps AccountAddress xm1 is used outputKey This is only relevant for the second type of processing that is when class SetChangeLog This defines the set of fields that are mapped to Key1 Key2 Key3 and Key4 fields of a ChangeLog record This can be different from the queryKey and primary key in order to match what is expected in ChangeLog DIS file records and also to support Change Log Summary processing If this parameter is not present then the primary key located in the given data type s data map file for example datamaps AccountAddress xm1 is used After these parameters are defined each data type contains a parameter for every field that is checked for changes If the value of the parameter is rze then changes ate tracked for the given field Otherwise changes are not tracked ChangeLog records are produced when the preprocessor is run for the given data type that is runDP sh AccountAddress A ChangeLog file is produced with a name prefix containing the data type that is ChangeLog AccountAddress In order to load these ChangeLog files the ChangeLog data loader is run that is runDL sh Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 117 Miscellaneous Util
139. Account Approval Pre Trade Approval uses Jurisdictions to limit user access to data in the database Records from the Oracle client that the Ingestion Manager loads must be identified with a jurisdiction users of the system must be associated with one or more jurisdictions In the Alert Case Management and Account Approval Pre Trade Approval system users can view only data associated with jurisdictions to which they have access You can use a jurisdiction to divide data in the database for example e Geographical Division of data based on geographical boundaries such as countries e Organizational Division of data based on different legal entities that compose the client s business e Other Combination of geographic and organizational definitions In addition itis client driven and can be customized In most scenarios a jurisdiction also implies a threshold that enables use of this data attribute to define separate threshold sets based on jurisdictions There can be two approaches to create a jurisdiction in the database e Creating Jurisdiction in the Database through Scripts e Creating Jurisdiction in Database through Excel Upload You can create jurisdiction in the database using the following steps 1 Add the appropriate record to the KDD_JRSDCN database table which Table 13 describes Table 13 KDD JRSDCN Table Attributes Column Name Description JRSDCN CD Code one to four characters that represe
140. Alert Correlation Rule Load Parameters Parameter Description load database username User to use to connect to the database when loading alert correlation rules DB_UTIL_USER load database password load conn driver Password for the above user This is set by the Password Manager Utility Database connection driver that the utility is to use oracle jdbc driver OracleDriver load conn url Database connection string that the Alert Correlation Rule Migration Utility is to use load schema mantas load system id load dirname MANTAS schema owner in the database in which loading of the alert correlation rule occurs MANTAS Name that is assigned to the system into which this instance of Alert Correlation Rule Migration loads meta data The system compares the value for this setting to the target system in the metadata file Directory from which the system loads alert correlation rule s XML files verify target system Check target name upon loading metadata files e Setting to N prevents Alert Correlation Rule Migration from checking the load system id against the target system specified when the alert correlation rule was extracted e Setting to Y enables this check If the target environ ment in the XML file does not match the setting for load system id or the target environment is present in XML file but the load system id is blank then the sys tem prompts you for an appropriat
141. Chapter 4 Data Ingestion Data Ingestion Flow Processes Data Ingestion Directory Structure Beginning Preprocessing and Loading The following sections take the high level view of Figure 12 and divide the Data Ingestion flow into distinct processes e Beginning Preprocessing and Loading Preprocessing Trading Compliance Solution Data e Processing Data through FDT and MDT e Running Trading Compliance Solution Data Loaders e Rebuilding and Analyzing Statistics e Populating Market and Business Data Tables e Processing Informatica Workflows and other Utilities The processes within each of the procedures refer to input and output directories within the Data Ingestion directory structure Where not called out in this chapter all Data Ingestion directories for example inbox or config reside in lt INSTALL_DIR gt ingestion_manager Also processing datestamps many Data Ingestion directories and subdirectories so that they appear with a YYYYMMDD notation The system provides this processing date to the set mantas date sh shell script when starting the first batch for the day Por detailed information about the Data Ingestion directory structure refer to section Data Ingestion Directory Structure on page 80 for more information In Figure 12 section A preprocessing begins The system executes preprocessors using the runDP sh script The following sample command shows invoking of a preprocessor INSTALL DIR inge
142. Correlation Rule Migration and other utilities require for processing Figure 65 provides sample information from the install cfg file that is specific to this utility Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 287 Alert Correlation Rule Migration Utility Chapter 8 Administrative Utilities HHHHHHHHHHHHHEHH CORRELATION RULE MIGRATION CONFIGURATION H H HHHHHHHHHHHHHHHEH HHHH GENERAL CORRELATION RULE MIGRATION SETTINGS Specify the name of the configuration file containing the names of correlation rules to be migrated This property is specific to the Correlation Rule Migration Utility corrRuleMig CorrRuleFileNm users mantas database db tools mantas cfg corrRule cfg EXTRACT These properties are shared by Correlation Rule Migration Utility with the Scenario Migration Utility Specify the database details for extraction extract database username utils database username extract database password utils database password Specify the jdbc driver details for connecting to the source database extract conn driver database driverName extract conn url jdbc oracle oci T209S8 Source System Id extract system id ENVIORNMENT Specify the schema names for Extract extract schema mantas schema mantas owner File Paths for Extract Specify the full path in which to place extracted Correlation Rules extract dirname users mantas database db_tools dat
143. DUMP DT 02 NOV 2006 Other Queries The Data Analysis Tool also supports providing SQL queries directly in the analysis XML file A query has two components the query title and the query itself As queries often contain characters that are reserved in XML you should follow the example below for escaping the SQL to ensure that it does not become corrupted QUERIES lt SQLQUERY title title select coll col2 from some table where some condition SQLQUERY lt QUERIES gt The following XML sample code illustrates use of the lt QUERIES gt element lt ANALYSIS gt lt QUERIES gt SQLQUERY title FO Transaction Roles gt lt CDATA select FOT mantas PRODUCT TYPE CD FOTPS PARTY ROLE CD count 1 as RoleCt from FO TRXN STAGE FOT FO TRXN PARTY STAGE FOTPS where FOT TRXN INTRL ID FOTPS TRXN INTRL ID group by FOT mantas PRODUCT TYPE CD FOTPS PARTY ROLE CD order by 1 2 gt lt SQLQUERY gt lt QUERIES gt lt ANALYSIS gt This code runs the query in the lt SQLQUERY gt element and writes the results to the output file For SQL queries the results are always in HTML Your code can contain any number of SQLQUERY elements The system runs each query in sequence after the other components of analysis are complete SQLQUERY Element Rules Several cautions and notes are specific to the lt SQLQUERY gt element e Ifyour query contains characters that XML standards reserve for example gt or lt you mu
144. Data Points Data Point Relationship Organization Root of an Oracle client s organization hierarchy Associated with 0 n users as a line organization Associated with 0 n users for view access to the organization Associated with 1 n Business Domains Associated with 1 n Scenario Groups Associated with 1 n Case Type Subtypes Associated with 1 n Jurisdictions Has no direct relati onship with a Role Role Associated with 0 n Users Has no direct relationship with an Organization User Associated with 1 Business Domains Associated with 1 n Jurisdictions Associated with 1 n Roles Associated with 1 n Scenario Groups Associated with 1 n Case Type Subtypes Associated with 1 n Organizations as members Associated with on e Organization as mantasLineOrgMember Users Admin Tools Should be mapped only to mantas Admin Role Scenario Class Associated to 0 n users Associated with Scenarios referenced in KDD SCNRO table Case Type Subtype Associated to O n users Group name identi fies the case type subtype matching a case CASE TYPE SUBTYPE CDinthe KDD CASE TYPE SUBTYPE table Business Domains Associated to O n users Business domain key must be in the KDD_BUS_ DMN table Jurisdiction Associated to 0 n users Jurisdict
145. Description This table is the definition of a report It identifies the name fre quency and any links to Active Pages applications Comments The Report Type Code should be AP for reports to be Oracle Financil Services6 1 3 outside of the Behavior Detection UI The Active Pages Applica tion Name should point to the Altio project name and the view should point to the path of the default view file to be loaded The Access Control Name should detail the acl required to access reports of this type Report Definition KDD REP DEFN PARAM This table contains a list of param Parameters eters that are valid for a particular type of report Report Template KDD REPORT TEMPLATE This table stores a template for a Please note that the Calen report This template is used to dar names that you specify generate report tracking records here must exists in the for different time periods KDD CAL table Also a report definition can have multiple templates associ ated with it for different organizations or calendars Report Template KDD REPORT TEMPLATE J Thistable stores jurisdictions to be Jurisdiction RSDCN used with a particular report tem plate Report Template KDD REPORT TEMPLATE P This table stores the values of The parameters and their Parameters ARAM parameters specific to the report corresponding values template would be set as default report parameters when the report is loaded th
146. E TRADE and TRADE EXECUTION EVENT Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 247 Data Retention Manager Chapter 7 Batch Processing Utilities Daily Partitioning Alternative Partition Structures Recommended Partition Maintenance Alternative Monthly Partition Maintenance 2 Drops the oldest partitions for all partition types 3 Inserts the saved data into what is now the oldest partition applicable to tables with open orders 4 Creates the new partitions 5 Recompiles the views that scenarios use The Data Retention Manager also enables you to build five daily partitions only a weckly basis rather than daily You do this by executing the run drm utility sh shell script and setting the drm weekly proc flg parameter to Y Refer to Table 95 on page 245 This procedure eliminates the need to perform frequent index maintenance Oracle Financial Services recommends doing this for large market tables This approach builds the daily partitions for the next week When creating the five daily partitions on a weekly basis the Data Retention Manager should be executed ptior to the end of the current week to create partitions for the next week Note You must set the WEEKLY ADD FL parameter in the KDD DR MAINT OPRTN table to Y so that the procedure works correctly For more information about this parameter Refer to Table 97 on page 249 for more information The structure
147. E2 CD count 1 from ACCT group by ACCT TYPE1 CD ACCT TYPE2 CD I select CUST TYPE CD count 1 from CUST where DATA DUMP DT 15 NOV 2006 group by CUST TYPE CD Null and Padded Space Null and padded space count analysis provides the number of occurrences for null Count Analysis values and padded spaces for a particular field in a table You perform this analysis by identifying the table and one or more columns of interest The null analysis feature has the following limitations e The feature is optional e The field identified for the specified table can be analyzed only once within the lt NULLS gt element per table e The filtering feature for the null analysis is optional and can have multiple constraints The structure to perform this analysis is ANALYSIS TABLES lt analysis for null counts occurs here gt lt TABLES gt lt ANALYSIS gt Within the lt TABLE gt element the name attribute identifies the table to be analyzed The targeted columns are identified within the lt NULLS gt element The field attribute in the lt NULL gt element sets each column name Apply filters to the analysis within the lt CONSTRAINT gt element The following code illustrates the structure for the a null and padded space count analysis lt TABLE name table name gt lt May contain one or more columns gt lt NULLS gt lt With no constraints gt lt NULL field column name gt lt Wit
148. FDT SequenceBatchSize Specifies the batch size when retrieving 10000 OrderState sequence ids for OrderState records FDT SequenceBatchSize Specifies the batch size when retrieving 1000 OrderEvent sequence IDs for OrderEvent records during end of day processing FDT SequenceBatchSize Order Specifies the batch size when retrieving 10000 sequence IDs for Order records FDT SequenceBatchSize Trade Specifies the batch size when retrieving 10000 sequence IDs for Trade records FDT SequenceBatchSize Specifies the batch size when retrieving 10000 Execution sequence IDs for Execution records FDT SequenceBatchSize Specifies the batch size when retrieving 10000 94 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Ingestion Directory Structure Chapter 4 Data Ingestion Table 32 Datalngest xml File Configuration Parameters Continued reported market sales price as an execution s expected print price when no comparable market sales occur during the order s marketable periods Property Name Description Example FDT MarketDataSource Value Specifies the source of market data Valid File values are File for file based access or RMI for access using an RMI server not recommended for performance reasons FDT CalculateDisplayability Specifies whether to calculate displayability FALSE Value states FDT ExplainableCancelCodes Specifies a comma separated list of Value expla
149. Financial Services Behavior Detection Platform 6 1 3 Administration Guide 259 Process to Deactivate Expired Alert Suppression Rules Chapter 7 Batch Processing Utilities Process to Deactivate Expired Alert Suppression Rules The following shell script should be executed in order to deactivate Alert Suppression Rules that have expired based on the current system date run upd suppression recs sh This sctipt should be run as the last step in batch processing just prior to ending the batch It is important that this script is run after post processing has been completed that is not before the Alert Suppression job is executed 260 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide CHAPTER 8 Administrative Utilities Oracle Financial Services Behavior Detection Platform 6 1 3 provides utilities that enable you to set up or modify a selection of database processes This chapter focuses on the following topics About Administrative Utilities Data Analysis Tool Get Dataset Query with Thresholds Utility Scenario Migration Utility Alert Correlation Rule Migration Utility Watch List Service Watch List Service Alert Processing Web Services Password Manager Utility About Administrative Utilities Several Behavior Detection database utilities that configure and perform system pre processing and post processing activities are not tied to the batch process cycle Data Analysis Tool Assists
150. Group Member DIS files being loaded with their respective runDP runDL processes The utility can be run by executing the command runNUAM sh Trade Blotter records are optionally created by the FDT and are loaded into the KDD TRADE BLOTTER and KDD TRADE BLOTTER ACTVY tables The FDT is configured by default to not create these records so it needs to be configured to do so The parameter FDT DeriveTradeBlotter value in the DataIngestCustom xml file should be set to rze to enable this functionality These records can be loaded after the FDT has been run by executing the command runDL sh TradeBlotter runDL sh TradeBlotterActivity After all scenarios and post processing jobs have been run an additional script needs to be run to score the trade blotter records based on the alerts that have been generated This process updates the KDD TRADE BLOTTER table and can be run by executing the command runScoreTradeBlotter sh 118 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Copying Thomson Reference Files Chapter 4 Data Ingestion Trusted Pair Refer to Score Trade Blotter on page 185 for more information The Trusted Pair DIS file is different from typical DIS files in that it is used to populate two separate tables KDD_TRUSTED_PAIR and KDD_TRUSTED_PAIR_MBR These tables can be populated by executing the commands runDP sh TrustedPair runDL sh TrustedPair runDL sh TrustedPairMember
151. H INF NDR NPR OHL HAI AIS MDS FeedUpCodes value Specifies status codes within the market data 0 that indicate that trading is active MDT Specifies properties used to configure the MDT component MDT NumberOfThreads value Specifies the number of worker threads that 4 the MDT uses when processing data MDT TickCodes Rising Specifies the tick code to use when the price 0 is rising MDT TickCodes SameRising Specifies the tick code to use when the price 1 is the same but the trend is rising MDT TickCodes Falling Specifies the tick code to use when the price 2 is falling MDT TickCodes SameFalling Specifies the tick code to use when the price 3 is the same but the trend is falling MDT MarketDataSource value Specifies the source of market data Valid File values are File for file based access or RMI for access using an RMI server not recommended for performance reasons MDT BufferSize value Allows an override to the BufferSize value property for MDT FDT Specifies properties used to configure the FDT component FDT NumberOfThreads Value Specifies the number of worker threads that 4 the FDT uses when processing data FDT LowerDisplayLimit Value Specifies the quantity below which orders are 100 exempt from display FDT UpperDisplayLimit Value Specifies the quantity above which orders 10000 are exempt from display FDT OrderPriceLimit Value Specifies the dollar value above which orders 200000 are exempt from display
152. ISPL NM Element ScenarioClassCode KDD REVIEW SCNRO CLASS CD Element OwnerUserName KDD REVIEW OWNER OWNER ID Element ClosingAction KDD REVIEW CLS ACTVY TYPE CD Element CloserUserName KDD REVIEW OWNER OWNER ID Element DateEnteredStatus KDD REVIEW STATUS DT Element AssociatedAccountServiceTeamID KDD REVIEW FINANCIAL SRVC TEAM INTRL ID Element AssociatedAccountRepresentativelD KDD REVIEW FINANCIAL RGSTD REP ID Element AssociatedAccountRepresentativeName KDD REVIEW FINANCIAL RGSTD REP NM Element AssociatedAccountRepresentativeLineOrganizatio KDD REVIEW FINANCIAL BRNCH INTRL ID nID Element AssociatedAccountRepresentativeLineOrganization KDD REVIEW FINANCIAL BRNCH NM Name Element AssociatedAccountRepresentativeSupervisoryOrga KDD REVIEW FINANCIAL SPRVSRY ORG INTRL nizationID ID Element AssociatedAccountRepresentativeSupervisoryOrga KDD REVIEW FINANCIAL SPRVSRY ORG NM nizationName Element AssociatedSecurityID KDD REVIEW FINANCIAL SCRTY INTRL ID Element AssociatedSecurityName KDD REVIEW FINANCIAL SCRTY DSPLY NM Element AssociatedTraderlD KDD REVIEW FINANCIAL TRDR INTRL ID Element AssociatedTraderName KDD REVIEW FINANCIAL TRDR DSPLY NM Element AssociatedInvestmentManagerlD KDD REVIEW FINANCIAL NVSMT MGR INTRL ID Element AssociatedInvestmentManagerName KDD REVIEW FINANCIAL NVSMT MGR DSPLY NM Element Actions ActionName KDD ACTIVITY TYPE CD ACTVY TYPE SHORT N M
153. Informatica Workflow Types Workflow Categories Workflow Processing The Oracle client s solution determines the required Informatica workflows or a subset thereof e AML Brokerage refer to AML Brokerage Workflows on page 135 for more information e AML Banking refer to AML Banking Informatica Workflows on page 146 for more information e Broker Compliance refer to Broker Compliance Informatica Workflows on page 153 for more information e Fraud Detection refer to Fraud Detection Informatica Workflows on page 158 for mote information e Insurance refer to Insurance Workflows on page 166 for more information Caution If you are running multiple solutions you must perform table comparisons to avoid running duplicate workflows Each workflow can include one ot more of the following categoties e Miscellaneous Optional e Mliscellaneous Pre Watch List Independent e Watch List e Miscellaneous Post Watch List e Summary e Balances and Positions Note The Informatica categories are not required for all solutions Refer to each section in the above list for a complete list of required categories This chapter provides the required Informatica workflows for deriving and aggregating data based on the solution Discussions of the workflows appear in the order that processing must execute them during data ingestion and include figures and tables that describe each workflow Sequence numbers that the accompanyin
154. Join on page 267 for more information e Simple join between two or more tables with filter restriction Refer to Simple Join with Filter Restriction on page 268 for more information e Join count of distinct values for specific column Refer to Join Counts Analysis on page 267 for more information The join count analysis is structured within the following elements lt ANALYSIS gt lt JOINS gt lt analysis for referential integrity here gt lt JOINS gt lt ANALYSIS gt Simple Join A join is set within the lt JOIN gt element To retrieve the join count between two or more tables the joins are identified within the lt MULTIJOIN gt element Within this lt MULTIJOIN gt element multiple lt JOIN gt elements can be set Because a join retrieves the join count between two or more tables lt LEFT gt and lt RIGHT gt elements are used to indicate the tables The lt LEFT gt element identifies the first table and its field using the table and column attributes The table and column attributes for the lt RIGHT gt element identify the second table and field The structure for a simple join count analysis is lt MULTIJOIN gt lt May contain more than one JOIN element gt lt JOIN gt Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 267 Data Analysis Tool Chapter 8 Administrative Utilities LEFT table table name column column gt RIGHT table
155. ML Brokerage Workflows Chapter 5 Informatica Workflows 20 ORION_ Production 0760 w_ph0760_ASM_to_APTxSM ORION_Production 0750 Informatica Balances and Positions Workflows AML Brokerage Informatica Balances and Positions workflows derive attributes that are useful in assessment of the financial status of an account customer or household Figure 33 illustrates Informatica Balances and Positions workflows for AML Brokerage Sequence 22 Figure 33 Balances and Positions Workflows AML Brokerage The application completes this workflow in one sequence processing of the workflows does not require data from external predecessors Table 50 describes the Balances and Positions workflows in Figure 33 Table 50 Balances and Positions Workflows AML Brokerage roduction Sequence Number Folder Workflow Number Workflow Name Predecessor 22 MLM Brokerage P 1030 w p1030 update cbps from deal Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 145 AML Banking Informatica Workflows Chapter 5 Informatica Workflows AML Banking Informatica Workflows The following sections describe the required Informatica workflows for deriving and aggregating data for the AML Banking solution e Informatica Miscellaneous Optional Workflows AML Banking e Informatica Miscellaneous Pre Watch List Workflows AML Banking e Informatica Watch List Wo
156. Manager manages these records by performing either an insert or update of the database with the contents of the first duplicate record The system inserts the record if a record is not currently in the database with the same business key The record updates the existing database record if one exists with the same business key The Ingestion Manager handles additional input records with the same business key by performing database updates Therefore the final version of the record reflects the values that the last duplicate record contains 62 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Process Flow Chapter 4 Data Ingestion Preprocessing The Ingestion Manager preprocesses market and trading data as procedures in the Trading Compliance following sections provide Solution Data 1 When Ingestion Manager satisfies dependencies from Group2 and preprocesses ot loads the data in Group3 it executes the runMDS sh script to process market data This script invokes the Market Data server which does the following Supports preprocessing of market data through the following mechanisms m Preprocessing of queue based equity market data from a market data I O stream for example Reuters through TIBCO m Support of input of market data in flat files Assigns sequence numbers to market data records Stores market data so that Firm Data Transformer FDT and Market Data Transformer MDT can retriev
157. N Common 3090 5 MLM Brokerage C 3160 w ph3160 write inst associations to Is ORION Common 3010 ommon MLM Brokerage Com mon 3041 MLM Brokerage Common 3120 6 ORION Common 3100 w ph3100 load staging fuzzy matches ORION Common 3020 ORION Common 3070 ORION Common 3090 6 ORION Common 3180 w ph3180 write Is to link tables ORION Common 3110 ORION Common 3140 MLM Brokerage Common 3160 7 ORION Common 3150 w ph3150 load staging and validate watch list ORION Common 3040 ORION Common 3070 ORION Common 3090 ORION_Common 3 091 ORION_Common 3100 MLM_Brokerage_Common 3120 MLM_Banking_Common 3130 8 ORION Common 3170 w ph3170 update staging list risk ORION Common 3150 9 ORION Common 3190 w ph3190 apply risk to nonacct entities ORION Common 3005 ORION Common 3170 9 ORION Common 3200 w ph3200 apply membership to entities ORION Common 3170 10 MLM Brokerage C 3171 w ph3171 update account customer risk ORION Common 3190 ommon ORION Common 3200 162 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Fraud Detection Informatica Workflows Chapter 5 Informatica Workflows 11 ORION Common 3191 w ph3191 apply risk to acct entities MLM Brokerage Common 3171 ORION Comm on 3190 12 ORION Common 3210 w ph3210 update fotps activity risk ORION Common 3191 12 MLM Brokerage C 3220 w ph3220 update bot activity risk ORION Common 3191 ommon 12 MLM Brokerage C 3230
158. Nz and Padded Space Count Analysis on page 266 for more information e Join Counts Analysis for referential integrity between two or more tables Refer to Join Counts Analysis on page 267 for more information Other Queries as configured Refer to Other Queries on page 270 for more information For both distinct value counts and null counts you can specify optional constraints The XML format for two of the files is identical For a join analysis the XML format uses a filter element that is similar to a constraint However you must specify the table name To specify a constraint use the CONSTRAINT element The lt CONSTRAINT gt element requires three attributes e Field Database field name to which the constraint applies e Value Value being compared Operator Operator used in the comparison Table 101 lists valid code operators Table 101 XML Code Operators XML Code Operator Comparison Operator GT gt LT lt EQ LTE lt GTE gt NEQ lt gt EMPTY Blank Character The following code sample illustrates the use of the lt CONSTRAINT gt element CONSTRAINT field DATA DUMP DT operator EQ value 15 NOV 2006 gt To include a constraint that filters out null columns use the EMPTY operator and set the value to is not null The following example illustrates the use of the EMPTY operator CONSTRAINT field DATA DUMP DT operator EMPTY valu
159. Operator Operand Operator Operand For example a BOTH SCORE CT gt 0 b BOTH SCORE_CT gt 0 AND FROM SCORE_CT TO SCORE_CT Note A space character is expected between each Operand and Operator combination The precedence of an operation may be depicted using a pair of parenthesis and yn The alert attributes provided should be a valid column name from the MANTAS schema tables KDD REVIEW and KDD REVIEW FINANCIAL e The CDATA section of the XML tag AlertCorrAttrOperations is validated as follows The Correlation Alert operation should be CORR The valid operators can be gt lt gt IN NOT IN AND and OR The expression can be nested to any arbitrary length provided that it confirms to a general syntax Operand Operator Operand Logical_Operator Operand Operator Operand Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 295 Watch List Service Chapter 8 Administrative Utilities For Example a CORR SCNRO ID gt 0 b CORR SCNRO ID gt 0 AND CORR SCNRO ID CORR SCNRO ID Note A space character is expected between each Operand and Operator combination m The precedence of an operation may be depicted using a pair of parenthesis id and m The Correlation Alert attributes provided should be a valid column name from the MANTAS schema tables KDD ALERT CORR and KDD ALERT CORR SCNRO Watch List Service Watch list web service enables yo
160. Oracle Financial Services Behavior Detection Platform Administration Guide Release 6 1 3 December 2013 ORACLE FINANCIAL SERVICES Oracle Financial Services Behavior Detection Platform Administration Guide Release 6 1 3 December 2013 Document Control Number 9MN12 0002 Document Number AG 12 OFS 0002 6 1 3 01 Oracle Financial Services Software Inc 1900 Oracle Way Reston VA 20190 Document Number AG 12 OFS 0002 6 1 3 01 Second Edition December 2013 Copyright 1996 2013 Oracle and or its affiliates All rights reserved Printed in U S A No part of this publication may be reproduced stored in a retrieval system or transmitted in any form or by any means electronic mechanical photocopying recording or otherwise without the prior written permission Trademarks Oracle is a registered trademark of Oracle Corporation and or its affiliates Other names may be trademarks of their respective owners Oracle Financial Services Software Inc 1900 Oracle Way Reston VA 20190 Phone 703 478 9000 Fax 703 318 6240 Internet www oracle com financialservices Contents List of PIQUIOS oia tn auct er s eee teas xv EASE OF Tal DIT xvii About this GUID E TT xxi Who Should Use this Guide C xxi SCOPE ORS Guilde veio entente ce perte perte ure T eti tee s Fate p C PUEDE REG e xxii How this Guides Otc ami Zed sist sis ironiaa CES TOM Ft d A e e e ERG RAA xxiii Wher
161. P i presentation of user information on the desktop Chapter 7 Batch Processing Utilities provides information about the database utilities related to the batch process Chapter 8 Administrative Utilities provides information about the database utilities that are independent of the batch process Appendix A Logging describes the logging feature Appendix B Oracle Financial Services Software Updates describes the application of software updates hotfixes and their impact on customization Appendix C Report Management describes the configuration and integration details of the Report Framework within user interface Appendix D Informatica Workflow Details lists how the Informatica Workflows are used within the application Appendix E Moving Oracle Financial Services System Environment explains the steps to move the system from one environment to another The Index provides an alphabetized cross reference list that helps you locate information quickly Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide xxiii Where to Find More Information About This Guide Where to Find More Information For more information about Oracle Financial Services Behavior Detection Platform refer to the following documents Oracle Financial Services Behavior Detection Platform Scenario Manager User Guide Oracle Financial Services Behavior Detection Platform Administration Tools User Guide Oracle Fin
162. ProfileTransaction 17 9005 AccountProfile 9001 9002 17 ORION Production 0190 w p0190 aggregate fotps to asms ORION Common 3210 ORION Production 0150 18 ORION Production 0310 w p0310 update asm for daily activity 9005 ORION Production 0190 19 MLM Banking Pro 0145 w ph0145 truncate CBPTxSM duction 19 MLM Banking Com 0480 w ph0480 aggregate asm to cbsm MLM Banking Common 0081 mon ORION Produc tion 0310 19 ORION Production 0630 w p0630 build ATxSD from ASMS ORION Production 0310 19 ORION Production 0660 w p0660 FOTPSR to AASD ORION_Production 0310 19 ORION_Production 0750 w ph0750 truncate APTxSM 19 MLM Brokerage P 0280 w p0280 create clog activity records ORION Production 0310 roduction 20 ORION Production 0760 w ph0760 ASM to APTxSM ORION Production 0750 21 MLM Banking Pro 0146 w ph0146 CBSM to CBPTxSM MLM Banking Production 0145 duction MLM Bank ing Common 0480 152 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Broker Compliance Informatica Workflows Chapter 5 Informatica Workflows Broker Compliance Informatica Workflows The following sections describe the Informatica workflows that are required for deriving and aggregating data for the Broker Compliance solution e Informatica Miscellaneous Optional Workflows Broker Compliance e Informatica Miscellaneous Pre Watch List Workflows Broker Compliance e Informatica Summary Workflows Broker Compliance
163. R ID column The Scenario Manager and Oracle login user IDs must be identical b User name in the USER NM column c Default user role in the USER ROLE CD column To use the Scenario Manager the user needs the MNR Data Miner database role The MNR database role is responsible for adjusting the pattern logic of existing scenarios and employs data mining techniques to create new patterns and scenarios d Flag of Y es or N o in the ACTV_FL column to specify whether the user is active A sample SQL insert statement is INSERT INTO KDD USER VALUES KDD MNR KDD MINER MNR Y FT To Grant a Database To grant a database role to the Scenario Manager KDD_MNR user follow the steps Role 1 Access the KDD_USER_ROLE table Table 17 defines the attributes in the KDD_USER_ROLE table Table 17 KDD USER ROLE Table Attributes Column Name Description USER ID User s login ID USER ROLE CD User s database role 2 Enter the following information into the table using an SQL script User login ID in the USER ID column User role MNR in the USER ROLE CD column A sample SQL insert statement is INSERT INTO KDD USER ROLE values KDD MNR MNR 46 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide About Changing Passwords for System Accounts Chapter 3 Security Configuration About Changing Passwords for System Accounts Througho
164. S start date null end date null alert status NW Base Working Directory required to put the temporary log from Database Server ap storedproc logdir tmp The common Path required to put the SQL files to execute commonSQLFilePath users mantast Solaris10 mantas58 b09 Ti5010S10 Iron 13080 WAS database db tools data HHHHHHHHH DATA RETENTION MANAGER CONFIGURATION 1HHHHHHHEBHBHBBHHBHHHBHIHE Set the Data Retention Manager input variables here HH drm_operation P drm partition type D drm_owner schema business owner drm object name A drm weekly proc fl1zN HHHHHHHHH Email Notification HH H H H HH HH H H H H HH H H H H H H H H HE H HE HE HE HE HE HE HE H HE H HE H H HH The following sections contain information on configuring email notification information If you wish to use Exchange you must purchase Java Exchange Connector obtain a license and the jec jar file The license file must be placed in the mantas_cfg file and the jec jar file must be copied to the db tools lib directory Then edit the file db tools bin run push email ksh uncomment the JEC JARS line HEHE HEHE HE HEHEHE HE HE HE PEHE HE HE HE HE HE HE HE HE HE PEHE HE HE HE HE HE HHH HE HE H HE H H RHEE HEHEHE HH HHH HH RH HH Currently only smtp smtps or exchange email type smtp Number of notifications that can run in parallel notification threads 4 Max number of active db connections utils database max_connections 4 Continue
165. SHLD ID SEQ 30000000 30000000 39999999 KDD TSHLD HIST HIST SEQ ID SEQ 30000000 30000000 39999999 KDD TSHLD SET TSHLD SET ID SEQ 30000000 30000000 39999999 In order to update your database tables with recommended values use SQLPLUS ora similar tool A sample SQL statement to update a set of sequence is UPDATE KDD COUNTER Set min value 10000000 max value 19999999 current value 10000000 where sequence name in DATASET ID SEQUENCE ATTR ID SEQUENCE PARAM SET ID SEQ PTTRN ID SEQ RULE ID SEQ SCNRO ID SEQ JOB ID SEQ TSHLD ID SEQ NTWRK ID SEQ LINK ID SEQ NODE ID SEQ LINK SUMMARY ID SEQ NTWRK DEFN ID SEQ TYPE ID SEQ TAB ID SEQ LINK TYPE SUMMARY ID SEQ TSHLD SET ID SEQ HIST SEQ ID SEQ AGMNT INSTN ID SEQ SCORE ID SEQ SCORE HIST SEQ ID SEQ Commit Repeat for each environment remembering to change the values for min max and current 286 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Alert Correlation Rule Migration Utility Chapter 8 Administrative Utilities Alert Correlation Rule Migration Utility Logs Use the Alert Correlation Rule Migration Utility to migrate correlation rules and associated audit trails between development environment and the production environment To provide a list of correlation rules you create a file listing the correlation rule
166. Sample KDD PRCSNG BATCH Table with Multiple Country Processing PRCSNG BATCH NM PRCSNG BATCH DSPLY NM PRCSNG ORDER EOD BATCH NM HK Hong Kong 1 HK LND London 2 LND NY New York 3 NY Starting a Batch Process Manually Since Hong Kong s markets open first this is the first batch You should run the Calendar Manager and Data Retention Manager at the start of the HK batch Upon setup of the batches Behavior Detection processing begins with the start mantas batch sh shell script The final step in a batch is calling the end mantas batch sh shell sctipt To start a batch manually follow these steps 1 Verify that the Behavior Detection database is operational tnsping database instance name gt 2 Verify that the INSTALL DIR database db tools mantas cfg install cfg configuration file contains the correct source database connection information 3 Access the directory where the shell script resides cd INSTALL DIR gt database db tools bin 4 Run the batch control shell script Start mantas batch sh batch name gt where batch name gt is the name of the batch This parameter is case sensitive If you enter an invalid batch name the utility terminates and logs a message that describes the error The error message appears on the console only if you have output to the console enabled in the lt INSTALL_DIR gt database db_tools mantas cfg categories cfg file Refer to Configuring Console Out
167. T CATEGORY 2 warning location syslog 4 Specify the full path to the external 10g4j configuration file log4j config file eWORKFLOW LOG4J CONFIG FILE Specify where a message should get logged for a category for which there is no location property listed above This is also the logging location of the default Oracle Financial Services category lunless otherwise specified above Note that if this property is not specified logging will go to the console log default location Specify the location directory path of the mantaslog if the mantaslog was chosen as the log output location anywhere above Logging will go to the console if mantaslog was selected and this property is not given a value log mantaslog location Specify the hostname of syslog if syslog was chosen as the log output location anywhere above Logging will go to the console if syslog was selected and this property is not given a value log syslog hostname Specify the hostname of the SMTP server if an e mail address was chosen as the log output location anywhere above Logging will go to the console if an e mail address was selected and this property is not given a value log smtp hostname Specify the maxfile size of a logfile before the log messages get rolled to a new file measured in bytes If this property is not specified the default of 10 MB will be used Figure 66 Sample Logging Configurat
168. TACENCU NT WIRETRXN VW P NETACENCU NT CUACEMAIL TMP NETACENCU NT CUACPHONE P NETACENCU NT CUACTAXID P NETACENCU NT JRNL P NETACENCU NT WIREACBENE TMP NETACENCU NT WIRETRXN 3 Execute the scenario job The product job template ID is 114698631 FR NetworkOfAcEn fA C steps P NETACENCU NT WIREACORIG Execute the link analysis network generation job The product job template ID To populate the temporary tables for FR NetworkOfAcEn fAC scenario follow these 1 Execute these refresh temporary table processes these commands can be run in parallel rur o oHdu HAWN AUN AN HAM AN HAM AWN HAM AN uu 9g DB TOO MP NETACENCU NT ACCTEMAIL B TOOI LS re x LS re LS re P_NETACENCU_N1 MP NETACENCU NT ACCTPHONE n z n h_temp_ U NT ACCTTAXID CUACADDR Fresh temp Fresh temp T ACCTPSWRD tabl Fresh temp table sh ACCTADDR VW Fresh temp table sh VW h_temp_table sh _ VW sh_temp_table sh VW table sh VW Fresh temp table sh VW table sh VW table sh VW table sh VW _temp_table sh le sh MP NETACENCU NT WIREACBENE VW tabl le sh MP NETACENCU NT WIREACORIG VW P NETACENCU NT ACCTADDR P NETACENCU NT ACCTEMAIL P NETACENCU NT ACCTPHONE TMP NETACENCU NT ACCTPSWRD TMP NETACENCU NT ACCTTAXID TMP NETACENCU NT CUACADDR P NETACENCU NT CUACEMAIL P NETACENCU NT CUACPHONE
169. Table 65 Table 66 Table 67 Table 68 Table 69 Run or Stop Scripts by Component oo seen 82 Environment Variable Desctiptions ssssseeeeeenenn 85 Application Configuration Files s ssessssssssssrsssreesssessreessreessreessrressreerressseeess 86 Datalngest properties File Configuration Parameters 87 Datalngest xml File Configuration Parameters see 88 Error File Signatures Output by Component sese 104 Files Output by Market Data Transformer sse 104 Files that Market Data Loaders Read and Process 105 Backed Up Files by Component sese 105 Output Files from TIB Spoolets sss 107 Log Files Output by Component eene 107 Miscellaneous Java based Uktilities ssssssseeeee 114 Ghange Log Parametets niente tineis i 117 Reference Files mio beet e SR EROR EEE 119 Fuzzy Name Matcher Utility Configuration Parameters 124 Dates used by Control D ata iocos esce beet Ei 129 Informatica Workflow Table Descriptions seeeee 135 Optional Workflows AML Brokerage eee 136 Pre Watch List Workflows AML Brokerage eee 137 Watch List Workflows AML Brokerage eee 139 Post Watch List Workflows AML Brokerage eee 142 Summary Workflows AML Brokerage eee 144 Balances and Positions Work
170. Workflows Broker Compliance sess 153 Workflows Broker Compliance 154 Summary Workflows Broker Compliance see 155 Balances and Positions Workflows Broker Compliance 157 Post Watch List Workflows Broker Compliance 157 Optional Workflows Fraud Detection sseeene 158 Pre Watch List Workflows Fraud Detection sss 159 Watch List Workflows Fraud Detection see 161 Post Watch List Workflows Fraud Detection sss 163 Summary Workflows Fraud Detection seee 164 Pre Watch List Workflows Insurance eeeeeeee 166 Informatica Watch List Workflows Insurance seeeeeee 167 Post Watch List Workflows Insurance eee 169 Summary Workflows Insutance eeeseeeeeeeen eterne 170 Managing Database Activities with Utilities sss 197 Sample install ctg File reete eoi 207 Sample Logging Information in the categories cfg File 209 Sample Log Routing Information seen 210 Sample KDD CAL HOLIDAY Table Loading Sctipt ss 212 Sample KDD CAL HOLIDAY Table Loading Sctipt sus 213 Configuration Information eene tnter 217 Configuring Batch Control Utility eee sse 225 Sample KDD PRCSNG BATCH HIST Table
171. XML files If you do not specify notarget or target name on the command line the system prompts you to supply a target environment on each extracted file Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 281 Scenario Migration Utility Chapter 8 Administrative Utilities To Extract Scenario To extract scenario dataset and network metadata follow these steps Metadata E 1 Navigate to the following directory cd INSTALL DIR 2 db tools 2 Edit the metadata configuration files with identifying information for the scenarios datasets or networks for extraction scnro ctlg id inthe scnros cfg file and or scnro ctlg id scnro id inthe scnros cfg file Note Providing both scnro ctlg id and scnro id in the scnros cfg file allows finer granularity when extracting scenarios If you provide both a scenario catalog ID and a scenario ID on a line you must sepatate them with a period data set id in the dataset cfg file network id inthe network cfg file 3 Execute the sm extract sh script in this order a Enter sm extract sh dataset to extract dataset metadata b Enter sm extract sh scenario to extract scenario metadata c Enter sm extract sh network to extract network metadata Loading Scenario The sm 1oad sh script loads translated XML table data files into the target database Metadata To Load Scenario To load scenatio dataset
172. a Specify the full path of the directory where the backups for the extracted scripts would be maintained extract backup dir users mantas database db_ tools data temp Continued on next page 288 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Alert Correlation Rule Migration Utility Chapter 8 Administrative Utilities Continued from previous page LOAD These properties are shared by Correlation Rule Migration Utility with the Scenario Migration Utility Specify the jdbc driver details for connecting to the target database load conn driver database driverName load conn url utils database urlName Target System ID load system id PROD ENVIRONMENT Specify the schema names for Load load schema mantas schema mantas owner Directory where scenario migration files reside for loading load dirname users mantas database db tools data Specify whether or not to verify the target environment on load verify target system Y Specify whether the Audit Trail History Records are to be loaded or not This property is specific to the Correlation Rule Migration Utility corrRuleMig loadHistory Y Specify the URL to be used for refreshing the Correlation Rules This property is specific to the Correlation Rule Migration Utility aps service url http 10 155 114 70 8080 mantas services AlertProcessingService Figure 65 Sample install cfgFile
173. a Data Miner or Data Analyst in determining how well a customer has populated the Production Data Model Refer to Data Analysis Tool on page 262 for more information Get Dataset Query with Thresholds Enables you to extract dataset SQL complete with substituted thresholds for analysis of the SQL outside of the Behavior Detection application Refer to Get Dataset Query with Thresholds Utility on page 274 for more information Scenario Migration Extracts scenarios datasets networks and associated metadata from a database to flat files and loads them into another environment Refer to Scenario Migration Utility on page 275 for more information Figure 53 on page 197 illustrates the frequency with which you use these database utilities Common Resources for Administrative Utilities Configuration files enable the utilities to share common resources such as database configuration directing output files and setting up logging activities For information about these resources and examples Refer to Chapter 7 Common Resources for Administrative Utilities on page 198 for more information Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 261 Data Analysis Tool Chapter 8 Administrative Ut ilities Data Analysis Tool The Data Analysis Tool enables you to determine how well a customer has populated the Production Data Model By reviewing the quality of data in each of the tables that th
174. a User Chapter 3 Security Configuration Setting up a User For accessing Account Approval Pre Trade Approval e Set of policies that associate functional roles with access to specific system functions e One or more associated organizational affiliations that control the user s access to Account Approval and Pre Trade Approval requests e Access to one or mote jurisdictions Access to one or more business domains For accessing Administration Tools e Setof policies that associate admin functional role with access to specific system functions To set up a user and provide the user access to the application perform the following steps 1 Create a user refer to the Oracle Financial Services Analytical Applications Infrastructure User Manual Release 7 3 for setting up a uset 2 Once the user is created map the user to the group This in turn maps the user to the role With this the user will have access to the privileges as per the role Refer to section User Group and User Roles on page 27 for more information on how to map User Roles to User Groups Note For the above sections Refer to the Oracle Financial Services Analytical Applications Infrastructure User Manual Release 7 3 for further information User Group and User Roles User Roles are predefined in the application Sample values for User groups are included in the installer but can be modified by clients to meet their specific needs The corresponding
175. a or b the user can view the record You can use this concept to protect special classes of data such as data about executives of the firm For example you can define a business domain as e Executives You can set this business domain with the employee account and customer records that belong to executives Thus only specific users of the system have access to these records If the executive s account is identified in the Private Client business domain as well any user who can view Private Client data can view the executive s record Hence it is important not to apply too many domains to one record 36 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide About Configuring Access Control Metadata Chapter 3 Security Configuration Creating Business Domain in the Database through Scripts Creating Business Domain in the Database through Excel Upload The system also stores business domains in the KDD_CENTRICITY table to control access to Research against different types of entities Derived External Entities and Addresses inherit the business domain set that is configured in KDD_CENTRICITY for those focus types There can be two approaches to creating a Business Domain in the database e Creating Business Domain in the Database through Scripts e Creating Business Domain in the Database through Excel Upload To create a business domain follow these steps 1 Add the appropriate use
176. able Description Multiple Files SSPARALLEL NUM Controls how many sessions run in parallel many workflows can split data into groups and run separate sessions in parallel Processing uses this parameter only for performance tuning SSJURISDICTION Controls the jurisdiction placed on derived External Entity and Address records prod bsm parm SSPERCENTDIFF Helps determine how much a security must move by the end of the day to be considered a win or loss If the security moves by less than a specified percentage processing does not count it either way If it moves by this percentage or more it counts as a win or a loss depending on whether the movement was beneficial to the account that made the trade prod mlm brokerage parm InactivityMonths Specifies the number of months that processing aggregated to determine whether an account is inactive If the sum oftrades and transactions over this number of months is lt 3 the account is considered inactive This setting can impact the Escalation in Inactive Accounts Scenario The default value is six months sMonthsRetention Controls how long Behavior Detection retains non standing instructions in the Instruction table prod_orion parm LookBackDays Specifies the following For sessions 720 727 set this param eter to 0 It controls how many days prior to today to examine for aggre gating wire activity between pairs of entities e For sessions 70
177. acle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 321 Informatica Workflows Appendix D Informatica Workflow Details Table 120 Informatica Workflow Description Continued Workflow Number Workflow Name Description 2172 w ph2172 update jurisdiction in ag This workflow updates the jurisdiction for ACCT GROUP 2180 w ph2180 instn identification This workflow creates unique identifiers for banks based on the third party vendors 2190 w ph2190 fotps instn processing This workflow marks all institutions with an Oracle Financial Services generated INTSN SEQ ID in FOTPS 2191 w ph2191 anticipatory profile inst This workflow marks all institutions with an Oracle n processing Financial Services generated INTSN SEQ IDin the Anticipatory Profile tables 2193 w ph2193 INS instn processing This workflow marks all institutions with an Oracle Financial Services generated INTSN SEQ ID in the insurance table 2200 w ph2200 build trxn tables from fo This workflow builds the records for all front office trans tps action tables 2220 w ph2220 update fot reversals This workflow adjusts the reversals for front office trans action tables 2450 w ph2450 populate ACCT SRVC TEAM This workflow maintains ACCT SRVC TEAM out of ACCT SRVC TEAM MEMBER 2600 w ph2600 loan smry mnth This workflo
178. advantages e Makes tables unavailable from the time the system truncates the data until it loads the data e Causes decreased performance if only a small per centage of records change on a daily basis e Makes multi batch processing unsuitable although the truncating process can be used for periodic or beginning of day table refresh Alternative Process Update Reference Data Mechanism that provides Behavior Detection with a set of records for updating in place in a table A look up of every record in the input determines whether Behavior Detection needs to update or insert data Advantages e Requires a minimal amount of updates for better per formance e Supports multiple batch processing and high data availability Disadvantages e Causes decreased performance if a large percentage of records require updating e Requires an Oracle client to determine what records have changed from day to day Single Instance Mechanism by which a single instance of Data Ingestion software processes an input data stream This configuration resides on one computer system Advantages e Makes processing easier to configure and monitor e Simplifies the Oracle client data extraction process Disadvantages Limits performance as a result of the limitations of a single computer system which results in a longer batch processing time Multiple Instances Mechanism by which configuration of multiple instances occurs on m
179. ager scripts runDL sh Order This command runs the data loaders for the order file that the FDT created previously The Ingestion Manager can execute the runDL sh scripts for trading and market data simultaneously For example Ingestion Manager can load ReportedMarketSale InsideQuote MarketCenterQuote and MarketState for market data simultaneously Figure 19 illustrates the Trading Compliance Solution data loading process Batch Loader Updates via Java Database Connection JDBC Data Inserts Figure 19 TCS Data Loading Process Rebuilding and When TCS market data loading is complete Ingestion Manager does the following Analyzing Statistics Figure 12 on page 55 1 Rebuilds database indexes by executing the runRebuildIndexes sh script for the loaded market data files refer to section Rebuilding Indexes on page 69 for mote information 2 Analyzes data table characteristics refer to section Analyzing Statistics on page 69 for more information The following sections describe these procedures 68 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Process Flow Chapter 4 Data Ingestion Rebuilding Indexes Analyzing Statistics Populating Market and Business Data Tables Processing Informatica Workflows and other Utilities During the data load process Ingestion Manager drops database indexes on some tables so that use of Oracle direct path loading can improv
180. al Roles eese essent tinte tinte tenente tenete tente tente tinens 46 Lo Granta Database Rolez ieteseceeee vore rip n ntn eer Ue Yap p n tru punte etian 46 About Changing Passwords for System Accounts sssssssssseeeeeeee tenente tenente neret tenente tenens 47 About Configuring File Type Extensions wissssssssssssssacsscscctsnstsosscscnsevssecushovanvevensvanvevensscsssnsssbessecesavensenceseveastvcestbeasevedse 49 CHAPTER 4 Data Ingestion oii tiii sant ea bso nese ener eoerenierereen 51 About Data Ingestioh x eet ee ener E EE E E E ER RE R ES 51 Process Blow sitcom a ees ER nep iot m US ERR 52 Data Ingestion Process Sumtnaty ee e dee e eee n i e M Aa et ies 54 Alternate Process Flow for MPIO Clients re rerit reete eere eei ER S ke EUER 56 Data Ingestion ElowPfOC6eSses tee te pier t e es D bes cii UA Tos eve etc G 57 Data Ingestion Directory SHUAUTEC sis tesss cases serat per teles p reri dte eer RU ee Ere PRESA 57 Beginning Preprocessing and LogahHps i e ret AAR t RAAR S 57 Preprocessing Trading Compliance Solution Data ssssssssssssssessssesrssssrsrsissrsissrrisrrissrrisssressressrressnrensrrensnrennrensnrensreness 63 Preprocessing Alternative 10 the MDS sa ies aa AEAEE 64 Processing Data through EDT asd MIT aie ioie eeii Ea EEEE EEEE E 64 Running Trading Compliance Solution Data Loaders eene tette tette tette 67 Rebuilding and Analyzing Statistics eee eee eene tette tette tete tete t
181. alerts for all remaining matches A alert is created for each of the remaining matches as long as they bind one of the centricity names in the bindings string This is the catch all algorithm that ensures that all matches that have a bound centricity value and a corresponding alert is created For a BIND MATCH grouping rule the system compares bindings KDD BREAK BINDING values for matches to determine whether it can group matches together into an alert For example the grouping algorithm interprets TRADER ASSOC SCRTY to create an alert each break set to be grouped must have a TRADER binding in which the values for that binding must match and each must either have an ASSOC_SCRTY binding in which the values match OR each must be missing the Assoc SCRTY binding Alerts that mentioned ASSOC_SCRTY could only be grouped with other alerts that mentioned ASSOC_SCRTY Similarly alerts that did not mention ASSOC_SCRTY could only be grouped with other alerts that did not mention ASSOC_SCRTY This list is order dependent and at least one binding should be marked as required using an exclamation point to prevent grouping of all miscellaneous matches into one big break The order helps determine the centricity in the first binding name in the binding string The centricity name is used to determine the alert s centricity ID Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 175 Update Alert Financial Data
182. alysis TOOL pe EI sesieiscptatcatavcseiaceasvasbsvescosucvcbsvabsavsunsdevcctustatacpustaassuutatssevstssshstasdpuvstiatasscuaddesbcusbdsealveblasvaesatuts 262 xii Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Contents Configuring the Data Analysis Tool sess eterne tenete tenens 262 To Configure General Tool Probertus eiie vete rer hp sts tess AAAA 262 ToConfeure the Analysts XME P3jes sso aree eta e PU e EE PE EE 263 Using the Data Analysis LOO bai o eret rete IR vote ne oe Ud x Ope teneor ER 271 To Ron the Data zAnalysis T00 1 ette etie dede aive eo RR dE 271 DEO GS cs cS MG 272 Understanding the Data Analysis Raport eee eet tte nentenentenees 272 Troubleshooting the Data Analysis Tool sse tenete ntn ete tenentes 273 Get Dataset Query with Thresholds Utility eese tette tente rnnn nene EIEEE Et 274 Using the Get Dataset Query With Thresholds Utility eese 274 Executing the Get Dataset Query with Thresholds Utility seen 274 To Execute the Get Dataset Query with Thresholds eee tenete tenete tentent 274 Scenario Miptation Utility E 275 Iioosscustotehctaunubon bur E iuam putei ub CNN ELM ate D DG C tecto rers 275 Using the Scenario Migration Utility eerte tenente tentato ntt torn torto ento issides 275 Configuring the Scen
183. ample size that the analyze procedure uses is 5 of the table under analysis To change the sample percentage update the SAMPLE PT column of the desired record in the KDD ANALYZE PARAM table IMP STATS Imports statistics that were previously calculated When running an ALL analysis the system exports statistics for the tables for later use Note Failure to run the statistics estimates can result in significant database performance degradation These scripts connect to the database using the user that the utils database username property specifies in the INSTALL DIR database db tools mantas cfg install cfg file The install cfg file also contains the following properties schema mantas owner schema market owner schema business owner The system derives schema names from these properties Note For Case Management Schema there is no separate script for managing Oracle database statistics But for improved query performance we have to manage the Oracle database statistics periodically Following are the sample commands To analyze table wise use use the following commands 252 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Flag Duplicate Alerts Utility Chapter 7 Batch Processing Utilities ANALYZE table lt Table name gt compute statistics Example ANALYZE table KDD_CASES compute statistics We can also perform whole schema analyze periodically Flag Duplic
184. an implement a distributed processing configuration that can run the Data Ingestion subsystem components on two or mote servers and let each server extract transform and load data for non overlapping data This distributed computing configuration influences configuration of Data Ingestion subsystem components or when to start and stop them The client is responsible for splitting data into non overlapping sets and placing this data into the inbox for each Data Ingestion instance For trading data and market data the client can split data by symbol ranges for example A through J on one server and K through Z on the other For reference data the client can select an arbitrary field to use in splitting the data Note that it is not necessary to split reference data and process it with multiple instances in situations where use of multiple instances processes trading and market data If ingestion of reference data occurs across multiple instances the client should ensure that ingestion of all reference data of a particular type occurs prior to ingesting data that is dependent on that type of data Data Rejection During Ingestion The Ingestion Manager can reject records at the Preprocessing Transformation or Loading stages The following sections provide an overview of the most frequent types of conditions that cause transactions to be rejected e Rejection During Preprocessing Stage describes how rejections occur during the Preprocessing s
185. ancial Publishing Global Bank Locations File tblabran txt The default Informatica source file directory must contain the rtsubase txt file Behavior Detection processes any changes The default Informatica source file directory must contain the tblabran txt file Behavior Detection processes any changes Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 119 Fuzzy Name Matcher Utility Chapter 4 Data Ingestion Copying Other Reference Files To copy reference files other than those that Table 41 contains use the following procedure To Copy Other To copy other reference files follow these steps Reference Files 1 1 Change to the directory identified by PMSourceFileDir for example software informatica pc7 1 4 MANTAS REP SrcFiles 2 Copy the ile country update file to this directory 3 If you do not have the reference files from Thomson Financial create empty files with the same names that is rtsubase txt and tblabran txt Fuzzy Name Matcher Utility During Informatica workflow processing the Fuzzy Name Matcher utility is used to match names of individuals and corporations candidates against a list of names targets The utility calculates a score that indicates how strongly the candidate name matches the target name All matches are case insensitive Although the system typically calls the utility from Informatica during Data Ingestion you can also execut
186. ancial Services Behavior Detection Platform FSDM Reference Guide Vol 1 Business Data Oracle Financial Services Behavior Detection Platform FSDM Reference Guide Vol 2 Oracle Financial Services Data Oracle Financial Services Behavior Detection Platform FSDM Reference Guide Vol 3 Case Management Data Oracle Financial Services Behavior Detection Platform Oracle Financial Services Service Guide Oracle Financial Services Behavior Detection Platform Configuration Guide Oracle Financial Services Analytical Applications Infrastructure DeFI User Manual doc Oracle Financial Services Analytical Applications Infrastructure User Manual Release 7 3 Oracle Financial Services Behavior Detection Platform Stage 1 Installation Guide Oracle Financial Services Analytical Applications Infrastructure Installation and Configuration Release 7 3 Oracle Financial Services Enterprise Case Management Installation Guide Stage 3 Por installation and configuration information about Altio software refer to the following documents AltioLive Deployment Guide AltioLive Developers Guide Por installation and configuration information about Sun Java System BEA and Apache software refer to the appropriate documentation that is available on the associated web sites xxiv Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Conventions Used in this Guide About This Guide Conventions Used in this Guide Table 1 list
187. and closed orders old executions old trades old derived trades lost order events and lost trade execution events logs yyyymmdd Status and error messages data errors Records that the system was unable to transform lt yyyymmdd gt data backup Backup of preprocessed input files lt yyyymmdd gt data firm load Transformed output files for loading into the database MDT Processing During execution of the runMDT sh script Ingestion Manager processes market data InsideQuote MarketCenterQuote and ReportedMarketSale through the MDT The Ingestion Manager also e Entiches data e Provides data for loading into FSDM schema The system executes the MDT with the runmDT sh script the following provides a sample command INSTALL DIR ingestion manager scripts runMDT sh 66 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Process Flow Chapter 4 Data Ingestion Figure 18 illustrates MDT data processing through runMDT sh Figure 18 Market Data Transformer MDT Processing When Ingestion Manager executes runMDT sh it places output files in the directories in Table 23 Table 23 runMDT sh Output Directories Directory Description data market transform Checkpoint data kept from one run to the next logs lt yyyymmdd gt Status and error messages data market load Transformed output files to be loaded into the database
188. and load data The workflow processes include the following e In Extract or Preprocess Workflow preprocessor components receive the raw market data business firm data and reference data from their external interfaces The components then perform data validation and prepare the data for further processing e In Transform Workflow transformer components receive the preprocessed market and business data The components then create derived attributes that support the downstream alert processing e In Load and Transform Workflow loader components receive preprocessed Reference data and transformed market and business data The components then load this data into the database In Post Data Load data transformations occur through Informatica derivations and aggregations risk assignment and watch list processing refer to Chapter 5 Informatica Workflows on page 133 for more information 52 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide p ng uoneusiuiupy e 9 uuoge d uonoejeg JorAeueg SEDIAJES JeroueuiJ 9 9219 es The following sections describe this process flow in more detail Extract Preprocess Transform Tibspool is a high performance C program that interfaces with live market feeds Market data can be read directly or requested via RMI Market Data ta i pM c tei ESA tem ee ed S Lx Sti GMT i y rocess Da oO le Trans f ile repr E
189. appender ref ref ConsoleAppender1 gt lt root gt Below mentioned logger levels are available DEBUG INFO WARN SEVERE FATAL W Application Server logs Logging levels can be configured in the below mentioned file lt SFIC_HOME gt conf RevLog4jConfig xml lt root gt lt priority value debug gt lt appender ref ref ConsoleAppender1 gt lt root gt Below mentioned logger levels are available Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 303 Logging Configuration File Appendix A Logging Sample Configuration File DEBUG INFO WARN SEVERE FATAL Services INSTALL DIR services server webapps mantas WEB INF classes conf mantas cfg install cfg INSTALL DIR services mantas cfg install cfg Ingestion Managet INSTALL DIR ingestion manager config install cfg The configuration file specifies enabling of priorities in a hierarchical fashion For example if Diagnostic ptiority is enabled Notice Warning and Fatal are also enabled but Trace is not In the configuration file you can specify the following Locations of recorded log messages Logging to the console files UNIX syslog e mail addresses and the Microsoft Windows Event Viewer Routing based on severity and or category Message library location Maximum log file size The following is a sample logging configuration file Make special note of the comments in the below sampl
190. apping a user to Multiple User Groups within Alert Management and Case Management Mapping a user to Multiple User Groups across Alert Management Case Management and Other Applications Mapping a Function to a Role AMACCESS CMACCESS RSGNTALL If a user is to have only one role then that user can be mapped to a single User Group associated with that User Role Refer to the Oracle Financial Services Analytical Applications Infrastructure User Manual Release 7 3 for more information about User to User Group mapping If a user needs to have more than one role within Behavior Detection then the user needs to be mapped to the different User Groups associated with the corresponding role When the user logs in user access permissions would be the union of access and permissions across all roles If a user needs to have different roles in both Alert and Case Management and roles for other platform supported applications then that user has to be mapped to different user groups When such a user logs in the user is taken to the Behavior Detection Start page rather than Behavior Detection Home page In the Behavior Detection Start page there are two menu links one for Alert Management and one for Case Management For any other platform applications the user is mapped to clicking each link opens the selected application in a new window The following list of functions need to be mapped to appropriate Alert and Case User Roles thro
191. application Once the application is installed you should execute these steps to avoid scenatio migration problems Make changes in only one environment 1 Only make changes to scenarios thresholds threshold sets and scoring rules in the source environment 2 Test and confirm your changes in the source environment 3 Extract scenarios from the source environment and migrate them to all of your target environments Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 283 Scenario Migration Utility Chapter 8 Adnministrative Utilities Conflicting sequence IDs are often the cause errors when you migrate a scenario so it is important to separate the sequence ID range Separate Sequence ID ranges 1 Review the MANTAS KDD COUNTER table which contains all sequence ID ranges and current values For information about the MANTAS KDD COUNTER table Refer to the Oracle Financial Services Financial Services Data Model Ref Guide Vol T 2 Start your sequence ID ranger at 10 000 000 and separate each environment by 10 000 000 The Oracle Financial Services Behavior Detection product sequence ID range is gt 100 000 000 Sequences to Modify You should set these sequences before doing any work on scenarios thresholds or scoring rules Table 110 Table 111 and Table 112 list sequences involved and sample values Table 110 Environment 1 Development
192. aration logical functions which include adapters for files and messages The functions also include mappings for data derivations and aggregations Behavior Detection Provides data access behavior detection and job services which include the Financial Services Data Model FSDM and scenatios specific to a particular solution set Alert Management Provides alert management reporting and searching of business data Case Management provides case management reporting and searching of business data Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 3 About the Oracle Financial Services Architecture Chapter 1 The Oracle Financial Services Behavior Detection Platform A set of components further divides each subsystem Components are units of a subsystem that can be installed separately onto a different server Table 2 outlines the definition for the subsystems and components In some cases however individual deployments can add subsystems or components to meet a client s custom requitements Table 2 Subsystems and their Components Common Name Administration Tools Directory Name admin_tools Contents Web enabled Administration Tools Case Management UI ftpshare lt case info Xmls for rendering the UI dom gt erwin forms Alert Management UI ftpshare lt alert info Xmls for rendering the UI dom gt erwin forms Case Management Web solution cm JSPs used i
193. ario Migration Utility iini miei iiri e foto tetto eerte riti pe 275 Extracting Scenario Metadata s een pp RU DP ERR E FEE RIEN DUREE 281 Loading Scenaro Metadata si tete tle e eere ee etes ele e d ER ehe ed oda 282 Sc nario Mistation Best Practice Ssenarini anA A mederi Debout 283 TROU coco Saws E T T A E E a e EA TEA 283 Best Dr chits anien tebe REE rn e e POR Haga Sot tt er deed b e e eet date eret RRA 283 Segunences to Modify ry insite 284 Alert Correlation Rule Migration Utility ccccssssssssssssisssseissesonssscatovsssosiavssennseedesesessveseibsedcbedesesesedevedossisoonedscvossevaseseee 287 TOG RT 287 Using the Alert Correlation Rule Migration Utility sssssssssse esent 287 Configuring the Alert Correlation Rules Migration Ufility eese esee eet tette tette tenente 287 Extracine Alert Correlati Hle sete gcc eset os dert E raqiea ettet te tentes nea dr Sawai 292 Leading Alert Correlation Rule 3a aet ei ipe epe re UE pedes 294 Watch List va 296 Alert Processing Web Services P 296 Instructions on Administering the Alert Processing Services sees 296 Password Manager Blityss se ccc ctt oret ere ee endete ng et eo d t tea a fee S 297 APPENDIX A EOOGQUAG E eM 299 About System Log Messages E
194. ase db tools mantas cfg install cfg configuration file job 300000 limit matches Y This produces the following e Matches 300000 19 e Alerts 300000 2 Actions 300000 2 To purge all matches in the first job in Table 81 regardless of alert affiliation set the following variables in the INSTALL DIR database db tools mantas cfg install cfg configuration file job 300000 limit matches N This produces the following e Matches 300000 22 300025 27 300030 32 e Alerts 300000 2 e Actions 300000 2 To purge only those matches that scenario 300001 generated between 11 06 2005 and 11 06 2006 with alert status OP set the following variables in the INSTALL DIR database db tools mantas cfg install cfg configuration file scenario 300001 Start date 06 Nov 2005 e e end date 06 Nov 2006 limit matches Y e alert status OP This produces the following results e Matches 300040 44 e Alerts 300007 Actions 300006 222 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Batch Control Utility Chapter 7 Batch Processing Utilities Batch Control Utility The Batch Control Utility enables you to manage and record the beginning and ending of an Behavior Detection batch process It also enables you to access the currently running batch You control the process through a job scheduling tool such as Maestro ot Unicenter Autosys This utility co
195. at is applied to alerts that satisty the operation This next step is either to close the alert or execute the Next operation Set NEXT OPRTN SET ID column or branch to further evaluate the alerts e heXPRSN ORDER ID column sets up an order of precedence by which the system attempts to satisfy the operations Note Enter NULL into the XPRSN ORDER ID column if that entry is linked from another entry that has a value in the XPRSN ORDER ID column TheALERT ATTR ID column identifies the attribute of the alert for evaluation e The OPRTR_CD column specifies the type of operation to be performed Allowed values ate gt gt lt contains or IN Note While using IN operator the right hand side variables should be separated by for example Nw OP e The value in the VALUE TX column provides the right hand side of the operation being evaluated e Ifthe current operation is satisfied and it is not the final operation in the operation set indicated by a NULL value in the NEXT OPRTN SET ID column the process jumps to the NEXT OPRTN SET ID If the NEXT OPRTN SET ID is NULL and the operation is true the system closes the alert e The DMN_cD column is the Oracle Financial Services application product code e hecLS ACTIVITY TYPE CD column specifies the activity type code of the closing action to associate with an alert that is closed by this rule Note The cLS ACTIVITY TYPE CD column is optional If the column is
196. ata Ingestion to alert post processing and enables access to the currently running batch Refer to Batch Control Utility on page 223 for more information Calendar Manager Updates calendars in the Oracle Financial Services FSDM system based on predefined business days holidays and days off or non business days Refer to Calendar Manager Utility on page 236 for more information Data Retention Manager Provides the capability to manage the processing of partitioned tables in Behavior Detection This utility purges data from the Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 195 About Administrative Utilities Chapter 7 Batch Processing Utilities system based on configurable retention period defined in database Refer to Data Retention Manager on page 241 for more information Database Statistics Management Manages statistics in the database Refer to Database Statistics Management on page 251 Flag Duplicate Alerts Enables you to run a script daily after the generation of alerts to identify pairs of alerts that are possible duplicates and adds a system comment to each alert Refer to Flag Duplicate Alerts Utility on page 253 for more information Notification Enables you to configure users of Alert Management and Case Management to receive UI notifications based upon actions taken on alerts or cases to which they are associated or when the alert or case is nearing a due
197. atch 23 highlight generation 184 186 system 23 Market Data Server 63 Market Data Transformer MDT 64 processing 66 Match Scoring 171 cloner executable 173 running 173 strategies 173 metadata access control 34 Alert Correlation Rule Migration 292 Scenario Migration 281 metadata files datasets 275 networks 275 scenarios 275 MiFID scenarios 56 missing data preprocessing 110 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 331 Index notification alert 172 Oo optional miscellaneous workflows 146 AML Banking 146 AML Brokerage 136 166 Broker Compliance 153 Fraud Detection 158 Oracle Financial Services accessing 26 Behavior Detection 8 hotfix 311 Ingestion Manager 8 jobs 11 17 19 20 21 22 24 logging 299 operations 7 post processing 171 software updates 311 Oracle Financial Services Architecture about 2 component view 2 3 deployment view 2 5 secutity view 2 Oracle Financial Services installer roles xxi Oracle Financial Services Job Protocol 11 12 Oracle Financial Services process 13 alert creation 9 alert scoring 9 assignment 9 auto close 9 auto suppression 9 172 highlight generation 9 historical data copy 9 match creation 8 match scoring 9 shell script 12 Oracle Financial Services processes netwotk creation 8 otganization usets 31 out of sequence events 111 output preprocessing 61 P Post Data L
198. ate Alerts Utility The Flag Duplicate Alerts Utility enables you to run a script daily after the generation of alerts This script identifies the pairs of alerts that are possible duplicates It then adds a system comment to each alert and identifies the paired alert in the comment as a Possible Duplicate External Entity focused scenarios in Behavior Detection can generate alerts either on external identifiers for example external account ID or on names of parties outside the bank The logic of the scenarios only generates the name focused alerts when the name has been found with multiple or no external identifiers This check is made across all transactions not just the transactions involved in a particular alert As a result a single run of an External Entity focused scenario can generate alerts involving the exact same transactions one alert focused on the external Party ID and one alert focused on the external Party Name Using the Flag Duplicate Alerts Utility The Flag Duplicate Alerts Utility looks at alerts that meet the following criteria e Entity focus EN e Status of New NW e Generated in the current running batch on the current date The utility selects and compares alerts that meet the listed criteria above It then determines whether generation of the alert is based on the same set of transactions for the same scenario with different focuses for example one alert is an ID and the other is a Name The utility flags t
199. ates the LAST RUN DT column in the KDD TA LAST RUN table values for RUN TYPE CD are ML I BC I and TC I with the date sysdate the procedure was last run run update ta utility sh This script calls the P TA MI UPDATE P TA BC UPDATE and P TA TC UPDATE procedures which update QLTY RTNG CD in the KDD TA ML DATA KDD TA BC DATA and KDD TA TC DATA tables respectively for any Review closed since the last run based on LAST RUN DT from KDD TA LAST RUN values for RUN TYPE CD ate ML U BC Uj and TC U The CLS CLASS CD value from KDD REVIEW is used as the new QLTY RTNG CD There are no parameters needed for this shell script It also updates the LAST RUN DT column in the KDD TA LAST RUN table values for RUN TYPE CD are ML U BC U and TC U with the date sysdate the procedure was last run The log for these scripts is written in the run stored procedure 1og file under the INSTALL DIR database db tools 1ogs directory Note The LAST RUN DT column in the KDD TA LAST RUN table is only updated for inserts and updates if at least one or more records were inserted or updated The LAST RUN DT column is not updated for significant errors that resulted in no records being updated These sctipts are a part of the database tools and reside in the INSTALL DIR database db tools bin directory You can run this utility anytime In other words it is not necessaty to run this utility during specific processing activities Oracle
200. ation gt gt User MYETERTSHETCIEGNIESETETHTSCTE USER Security Management System Security Attribute Administration TestOrgA iv Yes Yes pavan oracle com Choose User case anais 7 v GEN INST Business Domain Name GEN INST 5 Welcome CM Admin amp Thursday November 17 2011 Home Preferences arcis CICER User Administration Business Domain Description General Institutional Broker Dealer Heip Figure 9 Security Attribute Administration 3 Local intranet fa Using this UI an Administrator can map both Organizations and Users to different Security attributes 42 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Mapping Users To Access Control Metadata Chapter 3 Security Configuration Internet Explorer ORACLE Financial Services Enterprise Case Management Administration gt gt User Administration gt gt Security Attribute Administration Choose User Type User Pool Line Organization Parent Organization Own Case Flag Own Alert Flag Email Address Business Organization LE Organization 0 I5 Expand All Remove Jurisdiction gt Jurisdiction 0 9 Remove Business Domain Business Domain 2 Remove E Business Domain Code a b Scenario Class Scenario Cla
201. atterns will not be ignored This tom patterns N mode should be used when migrating scenarios between environments within the client s environment If a custom pattern is not in the loaded XML file then it will be deacti vated When set to Y any custom patterns will be ignored by the load process and should continue to operate load schema mantas MANTAS schema owner in the database in which loading of the scenario occurs MANTAS load schema business BUSINESS schema owner in the database in which load ing of the scenario occurs BUSINESS load schema market MARKET schema owner in the database in which loading of the scenario occurs MARKET load user miner DATA MINER schema owner in the database in which load ing of the scenario occurs KDD_MNR load miner password Password for the above user load threshold update Threshold values from the incoming scenario e Selecting N retains the threshold values from the tar get environment e Selecting Y updates thresholds in the target environ ment to values from the incoming file 280 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Scenario Migration Utility Chapter 8 Administrative Utilities Table 109 Scenario Load Parameters Continued Parameter Description load system id Name that is assigned to the system into which this instance of Scenario Migration loads metadata The sys tem
202. ault InputFileCharset DP Default SupplementalType Allows an override to the BufferSize value property for the XDP Specifies the character set of the DIS input files provided by the client Currently the only supported character sets are those that are compatible with ASCII Specifies an additional file type that a given XDP will create when it processes a file of the given type 100 UTF8 TrustedPairMember DP Data Type IndexField Specifies the field used to create an index into an input file Only valid for data types where AdapterClass is IndexFileAdapter SecurityIdentifier X DP EmployeeTradingRestricti on DescendOrgTree When processing EmployeeTradingRestriction records specifies whether to descend an organization s entire tree when creating records from an organization FALSE XDP Account DeriveAccountToP eerGroup When processing Account records specifies whether to derive an AccountToPeerGroup record if the AccountPeerGroupldentifier field is populated 98 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Ingestion Directory Structure Chapter 4 Data Ingestion Table 32 Datalngest xml File Configuration Parameters Continued Property Name Description Example on DescendOrgTree XDP EmployeeTradingRestricti If an Employee Trading Restriction record contains an Organization Identifier then it
203. avior Detection Platform 6 1 3 Administration Guide 109 Data Rejection During Ingestion Chapter 4 Data Ingestion Data Type Missing Data Referential Integrity Domain Values Every field in a record that processing submits to the Ingestion Manager must meet the data type and length requirements that the DIS specifies Otherwise the process rejects the entire record For example fields with a Date Type must appear in the format YYYYMMDD Thus the date April 30 2005 has a format of 20050430 and therefore is unacceptable In addition a field cannot contain more characters or digits than specified Thus if an Order Identifier in an Order record contains more than the maximum allowed length of 40 characters rejection of the entire record occurs The DIS defines fields that are mandatory conditional and optional If a record contains a field marked mandatoty and that field has a null value processing rejects the record For example all Trade Execution records must contain a Trade Execution Event Number If a field is marked conditional it must be provided in some cases Thus an Order record for a limit order must contain a Limit Price but an Order record for a market order need not contain a Limit Price In some cases you can configure Ingestion Manager to reject records that refer to a missing reference data record For example Ingestion Manager can reject an order that refets to a deal that does not appear in the Deal f
204. avior Detection Platform 6 1 3 Administration Guide 21 Performing Job Tasks Chapter 2 Behavior Detection Jobs Monitoring and Diagnosing Jobs In addition to the dispatch log file that records events for all jobs the system creates a job log for each job A job log records only the events that are applicable to that specific job By default a job log resides in the KDD PRODUCT HOME 10gs directoty You can configure the location of this log in the INSTALL DIR behavior detection algorithms MTS mantas cfg install cfg file If you do not know the location of the log directory check the insta11 cfg file The log mantaslog location property indicates the log location The default is KDD PRODUCT HOME 10gs but this location is configurable When troubleshooting a job processing problem first look at the file dispatch 1log for the sequence of events that occurred before and after errors resulted from a job Then look at the job log to diagnose the cause of the errors The job log provides detailed error information and clues that can help you determine why the job failed or generated errors The log file name for a job appears in the following format in the log directory job lt job_id gt lt date gt lt time gt log where job id is the job ID and date and time represent the job s starting timestamp If the job errors occurred due to a problem at the system level you may need to resolve it If you believe that
205. aximum look back or forward is 999 days calendar lookBack 400 calendar lookForward 14 HHEHHHHHHHHHHHH BATCH CONTROL CONFIGURATION H H H HHHHHHHHHHHHHHH When ending the batch age alerts in calendar or business days age alerts useBusinessDays Y Continued on next page Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 199 About Administrative Utilities Chapter 7 Batch Processing Utilities Continued from previous page THHHHHHHHHHHHHHE TRUNCATE MANAGER 1HHEHHBHEHBBHHBEBHHHBEHBHBBHBHHHBHHBUE Specify the database username and password for truncation manager truncate database username ingest user truncate database password ingest password HHHHHHHHHHHHHHHH SCENARIO MIGRATION CONFIGURATION 4HHH HH HH HH UE GENERAL SCENARIO MIGRATION SETTINGS Specify the flags for whether scoring rules and wrapper datasets need to be extracted or loaded score include N wrapper include N Specify the Use Code for the scenario Possible values are BRK or EXP load scnro use BRK Specify the full path of depfile and name of fixfile used for extraction and loading Note fixfile need not be specified in case of loading sm depfile users mantast Solaris10_mantas58 b09 Ti5010S10 Iron 13080 WAS database db tool s mantas cfg dep cfg sm release 6 1 3 HHH EXTRACT Specify the database details for extraction extract database username utils databas
206. ay be executed independently of each other Match Scoring Behavior Detection provides a mechanism to compute a score for matches to provide an initial prioritization Match Scoring rules are created using the Scoring Editor from the Administration Tools Refer to the Oracle Financial Services Behavior Detection Platform Administration Tools User Guide Release 6 1 3 for more information Running the Match Scoring Job The Match Scoring job is part of the Behavior Detection subsystem Behavior Detection delivers job template group 501 to run the Match Scoring job To Run the Match To run the Match Scoring job follow the steps Scoring Job 1 Verify that the dispatcher is running 2 Run the start_mantas sh lt template id gt script as follows start_mantas sh 501 All new matches in the system are scored Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 173 Alert Creation Chapter 6 Post Processing Tasks Alert Creation Matches are converted into alerts with the Alert Creator processes These processes are part of the Behavior Detection subsystem The system uses two types of Alert Creator jobs e Multi match Alert Creator generates alerts for matches that share a common focus are from scenarios in the same scenario class and possibly share other common attributes Each focus type has a separate job template e Single match Alert Creator generates one alert per match Running
207. base username property identifies in the INSTALL DIR gt database db tools mantas cfg install cfg file You run highlight generation after the creation of alerts and before the system ends the batch with the end mantas batch sh sctipt By default Behavior Detection writes log messages for this script in the INSTALL DIR database db tools logs highlights 1log file Augment Trade Blotter Oracle Financial Services Behavior Detection provides the ability to differentiate between various types of trades for example Client is Above 64 and Cancelled Trade using text based codes It also provides the ability to flag trades that require additional analysis before an analyst can mark trade as Reviewed or Reviewed with Follow up For this purpose the run augment trade blotter sh sctipt calls the P AUGMENT TRADE BLOTTER procedure which takes batch date as an optional input parameter If batch date is not specified the procedure operates on the current business date This procedure iterates through each trade and calls the P INSERT TRADE ATTRIBUTE and P UPDATE REQ ANALYSIS FL procedutes The database procedure P INSERT TRADE ATTRIBUTE contains the logic to assign characteristic codes to a trade It inserts data in the KDD TRADE ATTRIBUTE table The KDD TRADE ATTRIBUTE table contains the association between the trade TRADE SEQ ID and its characteristic text code ATTR TYPE CD The database procedure P UPDATE REQ ANALYSIS FL contain
208. base db tools mantas cfg install cfg file which Table 73 describes Table 73 HDC Configurable Parameters Recommended Parameter Value Descriptions hdc batchsize 10000 Number of break match key IDs are included in each batch thread for data retrieval hdc maxthreads 2x Maximum number of concurrent threads Number of CPUs that HDC uses for retrieving data to tune performance By default Behavior Detection writes log messages for this script in the INSTALL DIR database db tools logs hdc log file Alert Correlation Oracle Financial Services Behavior Detection provides a mechanism to correlate alerts to business entities and optionally to each other based on configurable rule sets This functionality is performed by the Alert Correlation process Details on configuring the data paths to correlate alerts to business entities as well as information on constructing the rules to correlate alerts to each other is provided in the following sub sections Running the Alert Correlation Job Alert Correlation is a part of the Behavior Detection subsystem The system delivers job template group 508 to run the Alert Correlation job for information on how to run this process though a web service Refer to the Oracle Financial Services Services Guide Release 6 1 3 To run an Alert Correlation job follow the steps 1 Verify that the dispatcher is running 2 Run the start mantas sh script as follows Start mantas sh
209. be taken on the last batch of the day Beginning of day processing Run Groups 1 through 5 sequentially Run AML Sian Data Informatica roel em parallel runDP lt data type gt Workflows runDL data type End AML Ingestion After running Group 2 refresh temp table sh Intra day processing repetitive C trading data ingestion process runDP sh lt data type gt TradeExecutian Order OpenOrder runMDS sh CorporateAction runMDT sh runFDT sh runDL sh data type runDL sh data type TradeExecution InsideQuote Order MarketCenterQuote OrderSummary ReportedMarketSale E xecution SecurityMarketState fade DerivedTrade CorporateAction runRebuildIndexes sh data type runDP sh InsideQuote InvestmentGuidelineOverride MarketCenterQuote ReportedMarketSale runDL sh InvestmentGuidelineOverride End of day processing runDP sh runDP sh SecurityMarketDaily SecurityFirmDaily runDL sh runDL sh SecurityMarketDaily SecurityFirmDaily market analyze sh firm analyze sh process market process firm summary sh summary sh End TCS Run BCS End Ingestion Informatica Workflows BCS ingestion Figure 12 Data Ingestion Process Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 55 Process Flow Chapter 4 Data Ingestion Alternate Process Flow for MiFID Clients Derivations done by the FDT process for the MiFID scenarios which use the Order Size Category requ
210. before awarding extra token match score The range is any integer gt 0 The recommended setting for corporations is 1 for personal names is 2 fuzzy name extra token pct decrease Determines the value of the extra token match score parameter in regard to extra tokens If multiple extra tokens are present reduction ofextra token match score occurs for each additional extra token The utility multiplies it by this number For example if extra token match score 50 and extra pct decrease is 50 percent the first extra token gets 50 percent the second extra token gets 25 percent the third token gets 12 5 percent the fourth 6 25 percent the fifth 3 125 percent etc The allowable range is O to 100 The recommended percentage for corporations is 100 percent for personal names 50 percent fuzzy name first first match score Allows the final score to be more heavily influenced by how well the first token of name 1 matches the first token of name 2 The allowable value is any real number gt 0 The recommended value for corporate names is 1 0 for personal names 0 0 fuzzy name match multi Determines how to handle multiple matches above the match threshold value If set to true the utility returns multiple matches If set to false it returns only the match with the highest Score fuzzy name file delimiter Specifies the delimiter character used to separate each columns in the
211. ble you to identify what jobs to run Table 4 provides an example of a job template group with two job templates Table 4 Kbb JOB TEMPLATE with Sample Job Template Group JOB ID TEMPLATE GROUP ID 37 1 41 1 12 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide About the Oracle Financial Services Job Protocol Chapter 2 Behavior Detection Jobs Understanding the Dispatcher Process The dispatcher process polls the job metadata waiting for jobs that need to be run To control system load the dispatcher also controls the number of jobs that run in parallel Generally the dispatcher process should be running continuously although it is possible to run jobs without a dispatcher For each job in the template group the dispatcher runs a mantas process The dispatcher tracks jobs for status and completion and reports any failure to the dispatch log Refer to Starting the Dispatcher on page 16 and Stopping the Dispatcher on page 17 for more information Understanding the mantas Process The dispatcher runs jobs using the mantas process This process runs the appropriate algorithm tracks status in the KDD_JOB and KDD RUN tables One mantas process can result in multiple KDD RUN records The mantas process also logs job progress and final status Applying a Dataset Override Configuring the Dataset Override Feature You use the dataset override feature to permi
212. bs that need to be run In between these checks the dispatcher sleeps for the time that you specify through the lt sleep time gt parameter in minutes Optional parameters include the following dispatch name Provides a unique name for each dispatcher when running multiple dispatchers on one machine JVM size Indicates the amount of memory to allocate to Java processing CAUTION For 32 bit Linux configurations Oracle Financial Services recommends running with the default JVM size 128 MB due to 2 GB process limit The script executes and ends quickly The dispatcher starts and continues to run in the background 16 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Performing Dispatcher Tasks Chapter 2 Behavior Detection Jobs Stopping the Dispatcher You do not normally shut down the dispatcher except for reasons such as the following e Problems while executing scenarios make it necessary to stop processing e The dispatcher and job processes are reporting errors e The dispatcher is not performing as expected e You must shut down the system for scheduled maintenance e You want to run the start_mantas sh restart_mantas sh Of recover_mantas sh script without the dispatcher already running You can then save your log files to the server on which you are working rather than the server running the dispatcher CAUTION If you shut down the dispatcher all active jobs sh
213. category Entries are placed in the configured default location in install cfg These can be uncommented and modified if routing by category is necessary log category ALERT PURGE location users orion mantas6 1 3 database db tools logs alert purge log log category BATCH CONTROL location users orion mantas6 1 3 database db tools logs batch control log log category CALENDAR MANAGER location users orion mantas6 1 3 database db tools logs calendar manager log log category DATA RETENTION MANAGER location users orion mantas6 1 3 database db tools logs DRM Utility log log category TRUNCATE MANAGER location users orion mantas6 1 3 database db tools logs truncate manager log log category COMMON UTILITIES location users orion mantas6 1 3 database db tools logs common utilities log log category EXTRACT location users orion mantas6 1 3 database db tools logs extract log log category LOAD location users orion mantas6 1 3 database db tools logs load log Continued on next page 208 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide About Administrative Utilities Chapter 7 Batch Processing Utilities Continued from previous page log category REFRESH TEMP TABI refresh temp table log LE location users orion mantas6 1 3 database db tools logs log category RUN STORED PROCEDURE location users orion mantas6 1 3 database db tools logs run stored procedure log lo
214. cenarios only post processing F Except OpenOrder and CorporateAction Rename ingest files in inbox e g filename _TWO_ dat y Extract copy DIS formatted data from third SRA party system for Intra day Batch n to inbox j Optional run firmanalyze sh marketanalyze sh Run detection for BEX and some TC scenarios only post processing Except OpenOrder and CorporateAction Rename ingest files in inbox e g filename NNN dat Run Intra day batch NNN Y Extract copy DIS formated data from 3 party system for Intra day batch n to inbox Only OpenOrder and CorporateAction Y Run End of day processes End TCM Ingestion 1 I I Figure 23 Intra Day Data Ingestion Processing Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 73 Intra Day Ingestion Processing Chapter 4 Data Ingestion Intra day processing references different processing groups as Figure 23 illustrates for example beginning of day processing and intra day processing as Figure 12 illustrates Multiple batches run throughout the day As in Figure 23 you configure batch ONE load and extract data and then start processing Data for OpenOrder and CorporateAction is not included When batch ONE processing is complete batch TWO processing begins The same occurs for all other batches until all batch processing is compl
215. ch the cases of a particular type and subtype can be grouped On categorizing the cases based on type and subtype they can further be grouped based on these subclasses Case Subclass 1 provides the list of subclasses for first level grouping Subclasses are not mandatory information for a case KDD_SUBCLASS2 Each record in the Case Subclass 2 table represents a subclass based on which the cases of a particular type and subtype can be grouped On categorizing the cases based on type and subtype they can further be grouped based on these subclasses Case Subclass 2 provides the list of subclasses for second level grouping Subclasses are not mandatory information for a case KDD TYPE CLASS MAP Each record in the Case Type and Class Map table represents the set of valid combinations of case type subtype subclass1 and subclass2 values which can be used to group the cases for proper investigation Refer to the Oracle Financial Services Behavior Detection Platform Financial Services Data Model Volume 3 Case Management fot mote information about these tables 38 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide About Configuring Access Control Metadata Chapter 3 Security Configuration Creating Case Type Subtype in Investigation Schema Adding Entries directly in the Table Using Script Adding Entries through Excel Upload You can create a Case Type Subtype in the investigation schema in the followin
216. cher should ignore the job and not process it This status identifies Job Templates The following status codes appear in the KDD JOB table when a job is processing e RUN tunning Implies that the job is running e FIN finished Indicates that the job finished without errors e ERR error Implies that the job terminated due to an error 18 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Performing Job Tasks Chapter 2 Behavior Detection Jobs Starting Jobs The Oracle Financial Services administrator starts jobs by running the start_mantas sh script To start a new job follow the steps 1 Create the new job and job description through an Oracle Financial Services Job Editor The application automatically assigns a unique ID to the job when it is created 2 Associate the new job to a Job Template Group using the KDD_JOB_TEMPLATE table Refer to section Understanding the Oracle Financial Services Job Protocol on page 12 for more information 3 Execute the start_mantas sh script as follows start_mantas sh lt template id gt The following events occur automatically 1 The job goes into the job queue 2 The dispatcher starts the job in turn invoking the mantas process and passing the job ID and the thread count to the mantas process 3 The mantas process creates the run entries in the metadata tables Each job consists of one or more runs 4 The mantas process hand
217. chments Note The export Web service stores the exported data in a configured location within your system For secutity putposes clients should restrict user access to the configured directory Directory Structure Each PDF is stored in a configured directory location Table 89 provides the directory structure for the exported data in INSTALL DIR database db tools Table 89 Directory Components Directory Components alert mgmt WEB INF classes conf mantas cfg install cfg Description Property to define the directory system where the exported data is stored in the absolute path investigation pdf export dir Child Directory Each PDF is exported to a child directory labeled with that alert or case iden tifier as well as the date and time Date and time of the batch action This is under the directory defined by investigation pdf export dir The directory name format Alert AlertID Date Time Case Case ID Date Time Date format YYYYMMDD Time format HH MM SS Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 231 Batch Export Utility Chapter 7 Batch Processing Utilities Table 89 Directory Components Continued Directory Components Child Directory Contents Description The PDF file name format Alert AlertID pdf Case CaseID pdf Alert ALERTID info xml Case CASEID info xml Document or spreads
218. cial Services Behavior Detection Platform 6 1 3 Administration Guide Scenario Migration Utility Chapter 8 Administrative Utilities Continued from previous page Controls whether jobs and thresholds are constrained to IDs in the product range product id range min through product id range max Values are Y and N If the range is not restricted you can use range check to fail the extract if there are values outside the product range extract product range only N extract product range check N LOAD load system id PROD_ENVIRONMENT Specify the jdbc driver details for connecting to the target database load conn driver database driverName load conn url utils database urlName Target System ID Specify the schema names for Load load schema mantas schema mantas owner load schema business schema business owner load schema market schema market owner load user miner utils miner user load miner password utils miner password load dirname Specify whether threshold can be updated load threshold update Y verify target system Figure 64 Sample install cfg File for Scenario Migration Note In the install cfg file entries are in the form Property1 Property2 That is the value for Property1 is the value that processing assigns to Property2 As such if you change Property2 s value Property1 s value also changes Configuring the To configure the envi
219. cle client provides The tibspool subdirectory contains files that the TibSpool component TIBS produces and the MDS treads These files are in a raw Tibco binary format and contain market data messages from a live market data feed TIBS writes files to the in subdirectory The MDS reads these files from the in subdirectory and moves the files to the backup subdirectory after extracting the data from them The in subdirectory within the tibspool subdirectory contains the raw data files that a TIBS instance extracts from the Market data feed During normal processing these data files reside in this location temporarily until deletion Each file name has the following format 106 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Ingestion Directory Structure Chapter 4 Data Ingestion tibspool backup Subdirectory logs Subdirectory tib lt A Z YYYYMMDDHHMMSS NNNNNN dat where lt A Z gt isa symbol range that corresponds to the symbol range that the TIBS instance is processing lt NNNNNN gt is a required sequence number to create unique file names Table 37 identifies sample output file names using the latter format Table 37 Output Files from TIB Spoolers Component Example Output Data Files TIBS tibSA Z 20011029112338 000001 dat tibSA Z 20011029112338 000002 dat The backup subdirectory within the tibspool subdirectory contains the backup of
220. column within the joined tables include the lt DISTINCT_COUNT gt element as content to the lt MULTIJOIN gt element The targeted table and its column are set to the table and column attributes respectively The following sample demonstrates integration of the DISTINCT COUNT element in the analysis MULTIJOIN lt JOIN gt lt LEFT table table name column column gt lt RIGHT table table name column column gt lt JOIN gt lt Optional Can only have one DISTINCT COUNT within the MULTIJOIN element DISTINCT COUNT table table name column column gt lt MULTIJOIN gt The lt DISTINCT_COUNT gt element is optional in the join analysis The following XML sample code illustrates use of the lt DISTINCT_COUNT gt element lt ANALYSIS gt lt JOINS gt lt MULTIJOIN gt lt JOIN gt LEFT table ACCT column ACCT INTRL ID gt RIGHT table CUST ACCT column ACCT INTRL ID gt Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 269 Data Analysis Tool Chapter 8 Administrative Utilities JOIN FILTER table ACCT column DATA DUMP DT operator EQ value 02 NOV 2006 gt DISTINCT COUNT table ACCT column ACCT TYPE CD gt MULTIJOIN lt JOINS gt lt ANALYSIS gt This sample code executes the following query select count DISTINCT a ACCT TYPE CD from ACCT a CUST ACCT b where a ACCT INTRL ID b ACCT INTRL ID and a DATA
221. compares the value for this setting to the target system in the metadata file load dirname verify target system Directory from which the system loads scenario network and dataset XML files Check target name upon loading metadata files e Setting to N prevents Scenario Migration from check ing the load system id against the target system spec ified when the scenario network or dataset was extracted e Setting to Y enables this check If the target in the XML file does not match the setting for load system id or the target is present in XML file but the load sys tem id is blank then the system prompts you for an appropriate action You can then continue with load or abandon the load and you can apply the same answer to all other files in the session of Scenario Migration or allow the utility to continue prompting on each XML file that has a mismatch Extracting Scenario Scenatio metadata includes XML files that contain the table data for scenario dataset Metadata and network logic The sm extract sh script invokes a Java tool which creates these files You start this script as follows sm extract sh mode notarget target lt name gt where mode mandatory is the scenario network or dataset notarget if included implies that the system does not save the target environment to the generated XML files target name identifies the same target in lt name gt for all extracted
222. cond scenario involved in the correlation adds to the score or NONE the score is not incremented above what has already been calculated Note The calculated correlation score is bounded by the values of the min_correlation_score and max_correlation_score parameters found in the following configuration files INSTALL DIR behavior detection algorithms mantas cfg install cfg for the Alert Correlation batch algorithm INSTALL DIR services install cfg for the Alert Correlation step of the PostAlert operation of the Alert Management Service CasePromotion optional Defines the parameters used to determine whether a newly discovered correlation should be promoted to a case Correlations that are already part of a case for example when a correlation is extended are not considered by this type of rule except the ExistingCasePromoteLossRcvryData element which determines whether of not to augment the existing case s fraud loss and recovery data with the related data from the new alerts added to the case Logical operations based on attributes of the correlation including scenarios involved in the correlation defined under AlertCorrAitrOperations can be used to determine whether or not the correlation should be promoted to a case The syntax supported operators and others are same as that of the A ertAttrOperations defined above including the CDATA requirement If the conditions result in the promotion of a correlation to a
223. cure FTP Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 311 Hotfix Effect on Customization Appendix B Oracle Financial Services Software Updates Hotfix Effect on Customization User Interface Scenarios When a hotfix is installed it can affect your customizations on the User Interface and Scenarios If your UI customizations are correctly isolated to the custom directory then the impact should be minimal It is possible however that the hotfix changes information in the base product that you have customized In that case you can not see the effect of the hotfix To minimize this be sure to avoid copying more than necessary to the custom directory For example you should not copy the entire BF Business xml file to override a few fields you should create a new file in the custom directory that only contains the fields you are overriding The hot fixes delivered will include installation and deployment instructions in the fix documentation If you have customized scenarios changed dataset logic or changed scenario logic then applying a hotfix to that scenario will remove those customizations If you customized datasets by creating a dataset override file Refer to Appling a Dataset Override on page 13 for more information then your custom dataset continues to be used after applying the hotfix It is possible that your custom dataset prevents the scenario fix from being evident i
224. d a range of actions taken on the alerts from either one job and associated scenario both jobs and their associated scenarios or the other job and scenario The sample Alert Purge Utility output explains the following situations e Sample Purge Alerts Utility Run Situation One shows how to purge those alerts that fully contain the first job in Table 81 Refer to section Sample Purge Alerts Utility Run Situation One on page 222 for more information e Sample Purge Alerts Utility Run Situation Two shows how to purge all matches in the first job in Table 81 regardless of their alert affiliation Refer to section Sample Purge Alerts Utility Run Situation Two on page 222 for more information Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 221 Alert Purge Utility Chapter 7 Batch Processing Utilities Sample Purge Alerts Utility Run Situation One Sample Purge Alerts Utility Run Situation Two Sample Purge Alerts Utility Run Situation Three e Sample Purge Alerts Utility Run Situation Three explains how to purge only those matches that are generated from scenario 300001 between 11 06 2005 and 11 06 2006 with status OP and ate wholly contained in alerts Refer to section Sample Purge Alerts Utility Run Situation Three on page 222 for more information To purge only those alerts that contain the first job in Table 81 set the following variables in the INSTALL DIR datab
225. d a threshold set output is Error Threshold Set not found Optional Redirect the output into a text file as follows INSTALL DIR database db tools bin get dataset query sh Data set ID Threshold Set ID query sql 274 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Scenario Migration Utility Chapter 8 Administrative Utilities Scenario Migration Utility Logs You use the Scenario Migration Utility to migrate scenarios datasets networks and associated metadata from the development environment to the production environment To provide a list of scenarios datasets or networks you edit the scnros cfg dataset cfg or the network cfg files prior to scenario extraction or loading The Scenario Migration Utility creates and migrates the following metadata files e Scenarios The scenario catalog identifier gt lt scenario id gt xml file contains scenario metadata for core Behavior Detection tables It also may contain scenario metadata for optional tables e Datasets The dataset idDS gt xml file contains dataset metadata for core Behavior Detection tables e Networks The network NW xml file contains network metadata for core Behavior Detection tables Note When the Scenatio Migration Utility extracts these files you can version control them or store them in the Oracle Financial Services client s atchival system To help avoid accidental loading of a sc
226. d on next page 202 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide About Administrative Utilities Chapter 7 Batch Processing Utilities Continued from previous page From address for sent mails This is ignored in Exchange mode If omitted in SMTP mode the mail account associated with the Unix Linux account is used email from SMTP settings email smtp host smtp port is usually 25 for smtp 465 for smtps email smtp port 25 email smtp auth false email smtp user email smtp password email smtp useHTML true Exchange settings See above for instructions to enable this Your Exchange administrator should help identify these settings email exchange server email exchange domain email exchange user email exchange password email exchange prefix Exchange email exchange mailbox email exchange useSSL true email exchange useFBA true email exchange useNTLM false email exchange draftsfoldername drafts email exchange useHTML true HTML email styles email le header font family Arial Helvetica sans serif font size 10pt color black email le hr color 555 background color 00 height 1px email le title font family Arial Helvetica sans serif font style bold size 12pt email le message font family Arial Helvetica sans serif font size llpt email style table font family Arial Helvetica sans serif b
227. d_asm_from_dump_file 0 BSM_Production 0110 w_p0110_reload_masm_from_dump_file 0 BSM_Production 0120 w_p0120_reload_nmsm_from_dump_file 0 BSM_Production 0130 w_p0130_reload_rrsm_from_dump_file Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 153 Broker Compliance Informatica Workflows Chapter 5 Informatica Workflows Informatica Miscellaneous Pre Watch List Workflows Broker Compliance Pre Watch List Informatica Miscellaneous workflows facilitate the application of customer supplied measures of risk to corresponding entities transactions and Sequence 1 instructions Figure 40 illustrates Informatica Miscellaneous Pre Watch List workflows for Broker Compliance Sequence 2 Sequence 3 Figure 40 Workflows Broker Compliance 2100 2180 2120 2190 Note Processing can occur simultaneously within each sequence Workflows with predecessors must wait for predecessor results before workflow processing can begin The application completes these workflows in three sequences processing of the wotkflows does not require data from external predecessors These workflows may proceed simultaneously in all sequences as soon as processing of any predecessors has completed Table 57 describes the Informatica Miscellaneous workflows in Figure 40 Table 57 Workflows Broker Compliance Sequence Number Folde
228. date and end date Figure 62 illustrates a sample table query an end time stamp in END TS and status of FIN in STATUS CD for the bolded entry indicates that the batch has ended PRCSNG BATCH ID PRCSNG BATCH NM DATA DUMP DT START TS END TS STATUS CD 1 DLY 10 Nov 06 11 Nov 06 11 Nov 06 FIN 2 DLY 11 Nov 06 12 Nov 06 12 Nov 06 FIN 3 DLY 12 Nov 06 13 Nov 06 13 Nov 06 FIN 4 DLY 13 Nov 06 14 Nov 06 14 Nov 06 FIN 5 DLY 14 Nov 06 15 Nov 06 15 Nov 06 FIN 6 DLY 15 Nov 06 16 Nov 06 16 Nov 06 FIN 7 DLY 16 Nov 06 17 Nov 06 17 Nov 06 FIN 8 DLY 17 Nov 06 18 Nov 06 18 Nov 06 FIN 9 DLY 18 Nov 06 19 Nov 06 19 Nov 06 FIN 10 DLY 19 Nov 06 20 Nov 06 20 Nov 06 FIN 11 DLY 20 Nov 06 21 Nov 06 21 Nov 06 FIN Figure 62 KDD PRSCNG BATCH HIST Table Batch End Status Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 229 Batch Export Utility Chapter 7 Batch Processing Utilities Identifying a Running Batch Process To Obtain a Batch Name Processing for Obtaining a Batch Name 3 Calculates the age of all open alerts and writes it to KDD REVIEW AGE if the EOD BATCH FL is Y in the KDD PRCSNG BATCH CONTROL table 4 Updates the KDD_REVIEW table for all alerts from the current batch to set the Processing Complete flag to Y This makes the alerts available for alert management 5 Deletes any records in the KDD Doc table that the system marks as temporary and are older than 2
229. de 333 Index suppression job alert 183 System 47 system crash 24 recovering jobs 24 T Technology Compatibility 1 Thomson Reference Files copying 119 identifiers 119 Trading Compliance data loaders 67 Transform Workflow 52 Truncate Manager 196 logs 258 using 258 U Update Alert Financial Data 171 user authorization 31 User Group and User Roles 27 user name 30 users access control 34 authorization model 31 case type subtype 31 jurisdiction 30 KDD_USER table 46 name 30 organization 31 roles 31 scenario group 31 using Calendar Manager Utility 237 utilities 195 261 Alert Correlation Rule Migration 287 Alert Correlation Rule Migration metadata 292 Alert Correlation Rules Migration 287 Alert Purge 195 215 annual activities 211 213 Batch Control 8 195 223 Calendar Manager 195 236 console output 210 Data Analysis Tool 261 262 271 272 273 Data Retention Manager 195 241 Database Statistics Management 196 251 Flag Duplicate Alerts 196 253 Flag Duplicate alerts 253 Get Dataset Query with Thresholds 261 274 highlight generation 184 Historical Data Copy utilities 185 Push E mail Notification 196 Refreshing Temporary Tables 196 255 Report Tracking 196 Scenario Migration 261 275 Scenario Migration metadata 281 Truncate Manager 196 Ww Watch List AML Banking 147 148 AML Brokerage 138 147 Fraud Detection 160 167 Watch List wo
230. ds they are m Compare fields on a single reference data record that can be identified by a primary key For example an Account record can be identified by an Account Identifier When an Account file is ingested the Primary Customer Identifier on Account XYZ is compared to the Primary Customer Identifier currently in the database for Account XYZ If they are different then a Change Log record is created This processing only accounts for updates to already existing records Change Log record is not created for new reference data records or deleted reference data records m Compare the set of values for a given field on several reference data records that map to a given key For example an Account Address record is identified with a combination of Account Identifier and Address Record Number However the information required is whether an Account Address record for a given Account has a field value that 1s different than any other Account Address record for that Account For example every Account Addtess record has a Country field Lets say there ate two Account Address records for Account XYZ in the database with values for Country of US and CN respectively On the next day an Account Address file is processed and there is an Account Address for Account XYZ with a value for Country of IR A Change Log record is generated for the Country field of 116 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide
231. duction 0 360 BSM_Production 0370 BSM_Prod uction 0380 17 9005 AccountProfile 9001 9002 18 ORION Production 0310 w p0310 update asm for daily activity 9005 ORION Production 0190 18 9006 HouseholdProfileDerived 9005 18 BSM Production 0431 w p0431 build MASD from MASMS BSM Production 0430 18 BSM Production 0460 w p0460 update masm for daily activity BSM_Production 0430 19 ORION Production 0630 w p0630 build ATxSD from ASMS ORION Production 0310 Informatica Balances and Positions Workflows Broker Compliance Informatica Balances and Positions workflows detive attributes that are useful in the assessment of the financial status of an account customer or household Figure 42 illustrates Informatica Balances and Positions workflows for Broker Compliance Solution 156 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Broker Compliance Informatica Workflows Chapter 5 Informatica Workflows Sequence 22 300 ioro i020 Sequence 23 Sequence 24 simultaneously within each sequence Sequence 26 Workflows with predecessors must wait for predecessor results before workflow processing can begin Sequence 25 soso Note Processing can occur Figure 42 Balances and Positions Workflows Broker Compliance The application completes these workflows in five sequences processing of the wotkflows does not require data from external predec
232. e e Alerts Near Due Date notifications Notification is generated to the owner of the alerts if the due date of the alert falls within the configurable parameter set in Installation parameter table The above notifications are generated after the complete execution of Batch provide the batch name and can be seen in the Notification Grid in landing page Each user sees the notification which is relevant to him 254 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Refreshing Temporary Tables Chapter 7 Batch Processing Utilities Note User can set the parameter of near due date and display of notification in KDD_INSTALL_PARAMS table Refer to the Configuration guide for more information Refreshing Temporary Tables Logs Some behavior detection patterns use the temporary tables as part of the detection process The log category REFRESH_TEMP_TABLE location property in the INSTALL DIR gt database db tools mantas cfg categories cfg file controls logging for this process The system writes log information for this process to the following location lt INSTALL_DIR gt database db_tools logs refresh_temp_table log Using Refreshing Temporary Tables The MINER schema defines these tables the tables have corresponding views that are used to populate them Prior to running these patterns run the refresh_temp_table sh script The script has the following calling signature refresh temp
233. e KDD CAL HOLIDAY table for the maximum date for each calendar in the table If the maximum date is less than 90 days ahead of the provided date the process logs a warning message that the specific calendar s future holidays need updating If any calendars have no holiday records the system logs a Warning message that the specific calendar has no recorded holidays for the appropriate date range Loading Non business Days After obtaining non business days or weekly off days typically Saturday and Sunday from an Oracle Financial Services client load this information for the upcoming calendar year into the KDD_CAL_WKLY_OFF table Figure 58 provides an example of a SQL script for loading the table INSERT INTO KDD CAL WKLY OFFS CLNDR NM DAY OF WK VALUES SYSCAL 1 INSERT INTO KDD CAL WKLY OFFS CLNDR NM DAY OF WK VALUES SYSCAL 7 COMMIT Figure 58 Sample KDD CAL HOLIDAY Table Loading Script Note By default the system identifies Saturdays and Sundays as non business days in the system calendar SYSCAL Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 213 About Annual Activities Chapter 7 Batch Processing Utilities Table 78 provides the contents of the KDD_CAL_WKLY_OFF table Table 78 KDD CAL WKLY OFF Table Contents Column Name Description CLNDR_NM Specific calendar name DAY OF WK Value that represents the day of the week
234. e The naming convention is to add x to the log files name for example mantas log mantas log 1 mantas log 2 so forth Note The log file size is a configurable property section Log File Sizes on page 307 provides instructions The default value for this property is 10 MB The maximum file size should not exceed two gigabytes 2000000000 bytes Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 299 Message Template Repository Appendix A Logging Message Template Repository The message template repository resides in a flat text file and contains messages in the format lt message id 1 gt lt message text gt The following is an example of a message repository s contents 111 Dataset id 0 is invalid 112 Run id 0 running Pattern 1 failed 113 Checkpoint false deleting match 111 112 and 113 represent message IDs whitespace and message text follow The O s and 1 s represent placeholders for code variable values Each subsystem has its own repository The naming convention for each message library file is mantas subsystem message lib lt language code gt dat whete subsystem is the name of the subsystem and lt language code gt is the two character Java ISO 639 language code For example the English version of the Algorithms message library is mantas algorithms message lib en dat The 1og message library property that the subsystem s base install cfg file
235. e functional areas Figure 31 illustrates Post Watch List Miscellaneous Informatica workflows for AML Brokerage External Predecessor 3130 sra Watch List Watch List Sequence 13 Sequence 14 Sequence 20 Note Processing can occur simultaneously within each sequence Sequence 21 Workflows with predecessors must wait for predecessor results before workflow processing can begin Figure 31 Post Watch List Workflows AML Brokerage The application completes these workflows in four sequences processing of wotkflows for one sequence requires data from an external predecessor Table 48 describes the Informatica Miscellaneous Post Watch List workflows in Figure 31 140 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide AML Brokerage Workflows Chapter 5 Informatica Workflows Note For the workflow _ph3501_Exp_and_Risk_Review_TP customer can configure the risk zones and customize the review reason text e Configuring Risk Zones The system compares a party s effective risk as it exists on the current version of the trusted pair member record with the party s current effective risk in the base table associated with the party ID type If the party s effective risk has increased by enough points to move it to a higher risk zone where risk zones are configurable ranges of risk values used to determine whether the increase in risk warrants a review of the pair the syste
236. e is not null gt You can also use the EMPTY operator to perform more complex comparisons than those that other operators support that Table 101 lists When using the EMPTY operator the generated SOL statement includes the field name a space and the text 264 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Analysis Tool Chapter 8 Administrative Utilities within the value string As such representation of more complex operations is possible An AND operator joins any existing multiple CONSTRAINT elements When adding date constraints as in the first example above you must specify the date in the same format as the databases NLS Date Format Oracle Financial Services recommends DD MON YYYY as the default format Distinct Values for Identifying the table and one or more column combinations of interest provides a Fields of Interest combination of distinct values and number of occurrences in the table The following Analysis code illustrates the required structure of this analysis within the following elements ANALYSIS TABLES analysis for distinct values occurs here gt lt TABLES gt lt ANALYSIS gt The name attribute of the lt TABLE gt element identifies the table against which this analysis is applied The vALUES element identifies targeted columns The field attribute of the lt COLUMN gt element sets each database column Application of filters to
237. e lt INSTALL_DIR gt database db_tools mantas_cfg install cfg configuration file contains the correct source database connection and logging information 3 Access the directory where the shell script resides cd INSTALL DIR gt database db tools bin 4 Start the alert purge shell script run alert purge sh Executing this command sets the environment classpath and starts the utility Upon execution of the run alert purge sh script the Alert Purge Utility generates a listing of actions matches and alerts that it needs to purge and records them in the INSTALL DIR database db tools logs Alert Purge 1log file The utility presumes that you have determined the input parameters to specify what matches alerts and actions to purge The utility does not check against the data to verify what it should purge Note To capture the SOL statements naming set log diagnostic true in the install cfg The parameters that define what matches to purge consist of one of two possible sets e A job ID which the KDD JOB table identifies Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 219 Alert Purge Utility Chapter 7 Batch Processing Utilities Automatic Restart Capability e A scenario ID as defined in the KDD_SCENARTIO table and a date range Behavior Detection does not support multiple scenario IDs so you should run them separately As part of this input set you can include an optiona
238. e 6 1 3 for more information Running the Alert Scoring Job To run an Alert Scoring Job follow the steps 1 Verify that the dispatcher is running 2 Run the start mantas sh script as follows Start mantas sh 504 where 504 is the job template that Oracle Financial Services provides to run the Alert Scoring algorithm 176 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Assignment Chapter 6 Post Processing Tasks Assignment Oracle Financial Services provides a mechanism to assign alerts and cases to a predefined owner either an individual user or a pool of users When performing alert or case assignment the module fetches new unowned alerts or cases for a given product and assigns them to an owner using a rule based strategy You can configure assignment rules by using the Administration Tools Refer to the Oracle Financial Services Behavior Detection Platform Administration Tools User Guide for more information Running the Assignment Job The Assignment Job is part of the Behavior Detection subsystem Oracle Financial Services provides default job templates and job template groups for running Assignment Job You can modify these jobs using the Administration Tools Refer to the Oracle Financial Services Behavior Detection Platform Administration Tools User Guide for more information To run an Assignment job follow the steps 1 Verify that the dispatcher is running 2 Run
239. e Financial Services PLOCSSING 7 Figure 5 Oracle Financial Services User Authorization Model 31 Figure 6 Sample SQL Script for Loading KDD JRSDON cence 35 Figure 7 Loading the KDD BUS DMN Table eee 37 Figure 8 Sample SQL Script for Loading KDD_ORG eee 41 Figure 9 Security Attribute Administration ssssssseeeeeeeee e ees 42 Figure 10 Components of Security Attribute 43 Figure 11 Data Ingestion Subsystem Components Workflow see 33 Ficure 12 Data Ingestion Process sitcom enel d 55 Figure 13 Dependency between process market summary sh and runFDT sh 56 Figure 14 Preprocessing Input and Output Directories sse 61 Figure 15 EGC Data Loading Process reete tede etes Ea 62 Figure 16 runMDS sh Input and Output Directories sse 64 Figure 17 Firm Data Transformer FDT Processing see 65 Figure 18 Market Data Transformer MDT Processing sse 67 Fipure 19 PCS Data Loading Process eroe empresas 68 Figure 20 Informatica Workflow Processing 70 Figure 21 Informatica Summary Generation ss sssssssssssriessressrteesrreesstssrtessrressrressresrreess 71 Figure 22 Informatica Watch List Processing sseeeeee 72 Figure 23 Intra Day Data Ingestion Processing seesseeeeeeeee 73 Figure 24 Data Ingestion Subsystem Directory Structure sss 80 Figure 25 Database A
240. e Financial Services Behavior Detection Platform 6 1 3 Administration Guide 147 AML Banking Informatica Workflows Chapter 5 Informatica Workflows Informatica Watch List Workflows AML Banking Informatica Watch List workflows facilitate the application of customer supplied measures of risk to corresponding entities transactions and instructions Figure 36 illustrates Informatica Watch List workflows for AML Banking External 3110 3160 2190 2180 3120 3091 3171 Predecessor Misc Misc Misc Misc Watch List Watch List Watch List Sequence 3 3000 3030 3005 Sequence 4 Sequence 5 UND Bm Sequence 7 3150 Sequence 8 3170 Sequence 9 3200 3190 Sequence 11 s Note Processing can occur simultaneously within each sequence Sequence 12 Workflows with predecessors must wait for predecessor results before workflow processing can begin Figure 36 Watch List Workflows AML Banking The application completes these workflows in nine sequences Processing of some workflows requires data from Miscellaneous predecessors refer to Informatica Miscellaneous Optional Workflows AML Banking on page 146 for more information These workflows may proceed simultaneously in all sequences as soon as processing of any predecessors has completed Table 53 describes the Informatica Watch List workflows in Figure 36 148 Oracle Financial Services Behavior Detection Platform 6 1 3
241. e MDT runMDT sh After receiving a soft kill the MDT terminates only when it has finished transforming all securities You can stop the MDT immediately by using the UNIX kill command to stop the process ID for the Java process that is a child of the runMDT sh script Note An MDT that you configure to run without polling that is batch mode stops automatically when no more data remains for processing However running a StopMDT sh script does not terminate batch processing runFDT sh Launches the FDT No stop script is available for FDT it stops after it processes all qualifying files that it finds in the data firm transform directory at the time the process starts The system processes an input file if the processing data and batch name are correct You can stop the FDT immediately by using the UNIX kill command to stop the process ID for the Java process that is a child of the runFDT sh process runDL sh data type Launches an instance of the data loader runDL sh You can configure the data loader to stop when it loads queued data for loading or to poll periodically until explicitly stopped with the stopDL sh script For example runDL sh Customer To run or stop a specific data loader specify a valid component that the run or stop script recognizes If the script does not recognize the component it exits with an error and identifies the valid list of parameters For valid component names refer to Figure 12 o
242. e Market Feed MDS QueueAdapter InputDir Specifies the directory that MDS should tibspool in examine when looking for live market data files that TibSpool produces MDS QueueAdapter BackupDir Specifies the directory to which MDS moves tibspool backup live market data files after processing them MDS QueueAdapter Specifies the file name prefix for live market tib InputFilePrefix data files that TibSpool creates MDS DataFeed Specifies the name of the exchange quote EXCHTIM ExchangeQuoteTimeFields time field in a live market feed MDS DataFeed Specifies the name of the market maker MarketMakerQuoteTimeFields quote time fields in a live market feed z DS DataFeed ICExchangeCodes 3 Specifies a set of mappings of RIC exchange codes to Behavior Detection exchange codes N XNYS B XBOS C XCIS P XPSE T XTHM PH XPHL A XASE MW XCHI Es zZ DS DataFeed Specifies a set of mappings of feed 1 XASE 2 XNYS 3 XBOS market data queries to account for out of order data that a live market feed provides FeedExchangeCodes exchange codes to Behavior Detection 4 XCIS 5 XPSE 6 XPHL exchange codes 7 XTHM 8 XCHI 43 XNAS MDS TimeInterval value Specifies the frequency in minutes with which 10 MDS writes output data files when processing data from a live market feed MDS CacheSize value Specifies the data cache byte size that the 1000000 MDS uses MDS RvSession Timeo
243. e _ARC archive tables The Platform UI displays alert information from the archive tables and information from the non archived tables This ensures that the alert information is in the same state as when the system generated the alert while the most recent information is available to the uset For mote information about the Platform user interface refer to the Oracle Financial Services Behavior Detection Platform User Guide Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 113 Miscellaneous Utilities Chapter 4 Data Ingestion Miscellaneous Utilities These utilities populate a single table in the data model They should be executed after all the files in Table 19 have been loaded with the following exceptions e AccountManagementStage utility should be executed before loading BackOfficeTransaction e Some utilities should run in between Informatica workflows Refer to Chapter 5 Informatica Workflows on page 133 for the sequence A utility should not be executed until all predecessors have executed successfully Commands to execute INSTALL INSTALL DIR ingestion manager scripts runUtility sh Utility Name DIR ingestion manager scripts runDL sh Utility Name gt These commands should be run serially The utility has executed successfully only after both of these commands have successfully executed A utility listed in the Predecessor column means that the g
244. e action You can then continue with load or abandon the load and you can apply the same answer to all other files in the session of Alert Correlation Rule Migration or allow the utility to continue prompting on each XML file that has a mismatch corrRuleMig loadHistory Load audit trail records on load e Setting to N prevents Alert Correlation Rule Migration Utility from loading the audit trail records from the XML file into the database e Setting to Y enables the system to load the audit trail records from the XML file into the database aps service url Web service URL of the AlertProcessing service to be used for refreshing the correlation rules Note Irrespective of whether the user specifies N or Y for the corrRuleMig loadHistory parameter a default audit trail record indicating the current load event is inserted into the database Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 291 Alert Correlation Rule Migration Utility Chapter 8 Administrative Utilities Extracting Alert Correlation Rule Alert correlation rule metadata includes XML files that contain the table data for the alert correlation rule along with their audit trails if any The sm extract sh script invokes a Java tool which creates these files To extract alert correlation rule metadata follow these steps jl Create a metadata configuration file someFileName cfg with identifying
245. e alert This is defined by the PRSDNC NB the lower the number the higher the precedence A NULL PRSDNC NB indicates not to apply this PATH ID to any alerts of this SCNRO ID or SCNRO CLASS CD Once alerts are correlated to business entities the alert to business entity relationships can be used to correlate alerts to each other Alerts will be grouped into a correlation if they share common business entities and if they meet the criteria defined in the Alert Correlation Rules These rules are managed through the Alert Correlation Rule Migration Utility see Art Correlation Rule Migration Utility on page 287 The logic of an Alert Correlation Rule is defined in XML and the Alert Correlation Rule Migration Utility is responsible for reading this XML from a file validating it and inserting it into 188 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Alert Correlation Chapter 6 Post Processing Tasks the KDD_CORR_RULE table The following is an example of the rule logic defined in an Alert Correlation Rule XML file followed by detailed descriptions of the elements contained in the XML file lt CorrelationRule id 123 name Possible Identity Theft gt lt MinAlertCount gt 2 lt MinAlertCount gt lt PrecedenceThreshold gt 5 lt PrecedenceThreshold gt lt AlertAttrOperations gt lt CDATA BOTH JRSDCN CD IN AMEA IND OR FROM SCORE CT TO SCORE CT gt lt AlertAttrOperat
246. e as they contain constraints that relate to properties and logging 304 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Logging Configuration File Appendix A Logging Specify which priorities are enabled in a hierarchical fashion i e if DIAGNOSTIC priority is enabled NOTICE WARN and FATAL are also enabled but TRACE is not Uncomment the desired log level to turn on appropriate level s Note DIAGNOSTIC logging is used to log database statements and will slow down performance Only turn on if you need to see the SQL statements being executed TRACE logging is used for debugging during development Also only turn on TRACE if needed log fatal true log warning true log notice true log diagnostic true log trace true H HHH HH H Specify whether logging for a particular level should be performed synchronously or asynchronously log fatal synchronous false log warning synchronous false log notice synchronous false log diagnostic synchronous false log trace synchronous true Specify the format of the log output Can be modified according to the format specifications at http logging apache org log4j docs api org apache log4j PatternLayout html NOTE Because of the nature of asynchronous logging detailed information class name line number etc cannot be obtained when logging asynchronously Therefore if this information is desired i
247. e into the database For reference data any file that has a load operation of Overwrite which the DIS specifies two options are available for loading data Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 61 Process Flow Chapter 4 Data Ingestion Guidelines for Duplicate Record Handling e Full Refresh Truncating of the entire table occurs before loading of data This mode is intended for use when a client provides a complete set of records daily e Delta Mode Updating of existing data and insertion of new data occur This mode is intended for use when a client provides only new or changed records daily The FullRefresh parameter in DataIngest xml controls the use of full refresh or delta mode When this parameter is rze the system uses full refresh mode when it is false the system uses delta mode Setting the default can be for either mode overriding the default for individual file types is also possible when needed The following sample command illustrates execution of data loaders INSTALL DIR ingestion manager scripts runDL sh Account Figure 15 illustrates the Trading Compliance Solution data loading process f iy Batch Loader a Updates via Java Database Connection JDBC Data Inserts Figure 15 TC Data Loading Process The Ingestion Manager considers records as duplicates if the primary business key for multiple records are the same The Ingestion
248. e it efficiently The Market Data server preprocesses market data files The following provides a sample command INSTALL DIR ingestion manager scripts runMDS sh This command initiates the Market Data server to process the ReportedMarket Sale InsideQuote and MarketCenterQuote files which the Oracle Financial Ser vices client previously placed in the inbox directoty 2 After Ingestion Manager preprocesses and loads the data in Group 2 it executes the runDP sh script to process TC trading data This script invokes Checking input trading data for errors Assigning sequence IDs to records Resolving cross references to market reference data Checking for missing fields or attributes When Ingestion Manager executes runMDS sh it places output files in the directories in Table 21 Table 21 runMDS shand runDP sh Output Directories Directory Description data market extract Market data intermediate files lt yyyymmdd gt logs lt yyyymmdd gt Preprocessing transformation and load status in individual date stamped log files data errors lt yyyymmdd gt Records that failed validation inbox lt yyyymmdd gt Backup of input files for restart purposes if necessary Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 63 Process Flow Chapter 4 Data Ingestion Figure 16 illustrates input and output directories for preprocessing market and t
249. e list contains 50 names the target name list contains 50 names and the fuzzy name max names per process property is set to 200 then one process will be run because the total number of names 100 does not exceed 200 If the candidate list contains 400 names the target name list contains 200 names and the fuzzy name max names per process property is set to 300 then four processes will be run each with 100 candidate names and 200 target names so that the max number of names per process never exceeds 300 The ability to break apart one large fuzzy matcher process into multiple processes through this property can help to overcome per process memory limitations imposed by certain Behavior Detection architectures fuzzy name max threads This parameter controls the number of threads to use when Fuzzy Name Matcher is being run Oracle Financial Services recommends that this value is not set to a number higher than the number of processing cores on the system fuzzy name max names per thread The purpose of this parameter is to keep the processing threads balanced so that they perform work throughout the course of the fuzzy matcher job That is instead of splitting the number of names to process evenly across the threads the value of this parameter can be set to a smaller batch size of names so that threads that finish ahead of others can keep working Oracle Financial Services Behavior Detection Platform 6 1 3 Admini
250. e load performance for high volume data After loading is complete Ingestion Manager rebuilds indexes using the runRebuildIndexes sh script which makes the table usable For example the system executes the runRebuildIndexes sh script on completion of the InsideQuote data loader INSTALL DIR ingestion manager scripts runRebuildIndexes sh InsideQuote The system then executes the firm analyze sh and market analyze sh scripts after rebuilding the indexes For example INSTALL DIR ingestion manager scripts firm analyze sh After rebuilding the indexes Ingestion Manager uses either the firm_analyze sh script for trading data or the market analyze sh script for market data to analyze data table characteristics This activity improves index performance and creates internal database statistics about the loaded data To build and update trade and market summary data in the database Ingestion Manager runs the process firm summary sh and process market summary sh scripts as in Figure 12 on page 55 The following examples illustrate execution of the scripts INSTALL DIR ingestion manager scripts process firm summary sh INSTALL DIR ingestion manager scripts process market summary sh After these two scripts complete processing Data Ingestion for Trading Compliance Solution is finished When the Data Ingestion processes finish loading data into the FSDM the Ingestion Manager performs the following tasks e Update
251. e schema identifies the Data Analysis Tool indicates how well the supplied data can support scenarios The tool does not make judgments about data quality Rather it provides a repeatable way to run a set of analytical queries across the data You can then use the results to direct further analysis The Data Analysis Tool e Counts all table rows in the schema e Identifies unique values and their distribution against the table e Determines the number of null occurrences for a specified column e Determines the number of padded spaces that occur for a specified column e Checks referential integrity between tables The following sections provide instructions for using the tool e Configuring the Data Analysis Tool e Using the Data Analysis Tool e Logs e Troubleshooting the Data Analysis Tool The tool provides its results in either a text or Hypertext Markup Language HTML file You can then use these results to direct an investigation for data quality Note To use the Data Analysis Tool effectively you must have basic knowledge of Structured Query Language SQL and Extensible Markup Language XML Configuring the Data Analysis Tool To Configure General Tool Properties The Data Analysis Tool uses the install cfg and analysis xml or similat configuration files You edit either file in a text editor such as vi To produce well formed XML files however you should edit the XML file in a validating XML editor Behavior Detec
252. e the utility through a UNIX shell script at the command line The Fuzzy Name Matcher Utility places its output in a delimited results file one match per line the utility overwrites a previous results file if one exists The output in the file includes e Match score normalized to a number between 0 and 100 e Delimiter e Candidate name e Delimiter e Matched name on the target list The results file contains all matches that are above a match threshold that is a single candidate name can match multiple names on the target list Candidate names that do not match any target names do not appear in the file Logs As the utility performs the name matching process it generates a log that it enters in the INSTALL DIR ingestion manager fuzzy match log Utilities log file the logging process time stamps all entries The log file contains relevant error and warning information Refer to Appendix A Logging on page 299 for more information about these configurable files You can modify the default logging configuration for this utility in the Ssoftware informatica mantas ingestion manager fuzzy match mantas cfg install cfg configuration file Figure 26 on page 121 provides a sample file 120 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Fuzzy Name Matcher Utility Chapter 4 Data Ingestion Log Configuration Items Hoe ce eae et ee Ba ee ENE Ae eee eL ee ied Specify which pr
253. e to Pind More Information tient treten tette rtn deterius titio eee vate oko ESEESE SEEE en XXIV Conventions Used tn this Guide iciuiscsisseseiniesssivessessasssasssicasbeteasisvaievedsanieieteasindasssotoensnseiavesssvotiededeacetia iotsspb vesueesivebe sets XXV CHAPTER 1 The Oracle Financial Services Behavior Detection Platform 1 Technology Compatibility scissessscstassscsasscssssseissstesedsscdossiesedassnceassodessseasssscauseatosvessdesssebsJotesusesessieasdssedubvsnesdosadesdendedetsseuesested 1 About the Oracle Financial Services Architecture osse retten NA AE RN AARE 2 Component VIeW cens cd EEA esconde E c e E ae cT aT cosines 2 Terga dae o OO E EE a pr Gh depo dO PE Pb RF d a WO PU REO ipiis 2 SUVE i EAE EREE E E E un bette AAE ATE RIT E AE ACRES EOE REE 3 Depl ym nt Vew s maoistai Ea E E tuno hat E E N a sp ret A e Beo bend 5 SECULIEY i d EEA R AANE EERE RE EEE ERE E AE A EEE RN AEAEE EEES EE ETER 6 About Oracle Financial Services Operations scscscsesesssessvoestesssenteisseseissodessoesesoisseoeiesesoessvoeiesoesssseisvesssvogioboassnoaiesotssnetes 7 iriudirir TM M 8 Data LAGES UO Merere naa eE oA eveniet aree E b e ect b e e ic AM 8 Behayiot Detection ettet i te ot recae ebbe t de t i be Ut D ee bie ocu Ee io cuotas 8 POoSEPtOCOSSIBO 2 cte eee e e e dace edet AT res ke bee bn 9 End Batch uote ure toot UH eater E O VPE D DR DER IT ORE Hee t OH Qe I LIRE 10 CHAPTER 2 Behavior Detection Job
254. e username extract database password utils database password Specify the jdbc driver details for connecting to the source database extract conn driver database driverName extract conn url jdbc oracle oci Ti5010S10 Source System Id extract system id Specify the schema names for Extract extract schema mantas schema mantas owner Continued on next page 200 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide About Administrative Utilities Chapter 7 Batch Processing Utilities Continued from previous page extract schema business schema business owner extract schema market schema market owner extract user miner load user miner extract miner password utils miner password File Paths for Extract Specify the full path in which to place extracted scenarios extract dirname users mantast Solaris10_ mantas58 b09 Ti5010S10 Iron 13080 WAS database db _tools data Specify the full path of the directory where the backups for the extracted scripts would be maintained extract backup dir users mantast Solaris10_ mantas58 b09 Ti5010S10 Iron 13080 WAS database db tools data temp Controls whether jobs and thresholds are constrained to IDs in the product range product id range min through product id range max Values are Y and N If the range is not restriced you can use range check to fail the extract if there are values
255. eb server logs mantas reveleus web context logs UMMService log Application server logs The following file is the message library for the Application Server logs FIC APP HOME gt common ficserver logs revappserver log Database objects logs DB objects logs used in the application are maintained in the table KDD_LOGS_MSGS An entry in this table represents the timestamp stage error code and module Ingestion Manager The following file is the message library for the Ingestion Manager INSTALL DIR ingestion manager config message dat 302 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Logging Configuration File Appendix A Logging Logging Configuration File You can configure common logging through the following files depending on the subsystem you want to modify Administration Tools lt FIC_HOME gt AM admin_tools WEB INF classes conf mantas_cfg inst all cfg Database INSTALL DIR database db tools mantas cfg install cfg Scenario Manager INSTALLED DIRECTORY behavior detection toolkit mantas cfg install cfg Behavior Detection INSTALL DIR behavior detection algorithms MTS mantas cfg install cfg Alert Management Case Management W Web Server logs Logging levels can be configured in the below mentioned file mantas reveleus web home gt conf RevLog4jConfig xml In below mentioned tag root lt priority value debug gt lt
256. ed from A to B TP REVIEW REASON TX PARTY2 and risk of lt Party2 gt increased from C to lt D gt The string for Review Reason Text parameters is translatable You can change these strings except the values in angular brackets like lt Party1 gt A lt B gt lt Party2 gt lt C gt and D Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 141 AML Brokerage Workflows Chapter 5 Informatica Workflows Table 48 Post Watch List Workflows AML Brokerage If the system determines that the Trusted Pair record that has experienced a threshold triggering risk increase is still in a Risk Escalated Recommend Cancel RRC state that is a Supervisor has not reviewed the recommendation the system appends the new review reason text to the existing reason text on the current Recommend Cancel version of the Trusted Pair record A semi colon and a single space is used as the method of appending Note While appending a new review reason text to the existing text the system finds that appending text will result in the field exceeding 2500 characters In this case the system will overwrite the existing review reason text on the current Rec Cancel version of the Trusted Pair record with the current review reason text The above mentioned parameters for configuring risk gones and customizing review reason text ate located in the prod_orion parm file under the section S_m_p
257. ee D P RR Eee n Rr hin PL ten re 30 Defining the User Access Properties and Relationships 30 Obtaining Information Before Configuring Access Control 33 About Configuring Access Control Metadata ettet tinte testet ntpote ties sbesesoasossensooniesesseneees 34 Creating Jurisdiction m the Database e eet eese iret tete hus e E E 35 Creating Jurisdiction in the Database through Seribts eese eene tette tnnt nte nennen nnns 45 Creating Jurisdiction in Database through Excel Upkoad eese netten tentent 36 Creating Business Domain in the Database tenete nennentnnennetntetntentntens 36 Creating Business Domain in the Database through Scripts eee eese 37 Creating Business Domain in the Database through Excel Upload eene 37 Creatine Case Tybe SUIS us ester e eoi brake kic ds Guo tos TU condom RE B cU A ate A E em Ru pil RU 38 Creating Case Iype Subtype in Investigation SC eid ume ete niet q erpresicta cach pedestal er e bride 39 Adding Entries directly in the Table Using Seribt sees eene tentent tette tenentes 49 vi Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Contents Adding Entries through Excel Ubload uu eat temere arie re ko terti Pe o oH eek 49 Creating Case Subclass1 in Investigation Schema sse nene nntetnterntnntnnens 39 Adding Entries through Excel Upload esee aen nee eee qae 49 Creating Case Subclass2 in Inv
258. efer to Assignment on page 177 for more information Account Approval Pre Trade Approval Tasks Two processes Load Employee User Groups and Request Assignment are required to successfully utilize the Account Approval AA and Pre Trade Approval PTA application Refer to Account Approval Pre Trade Approval Tasks on page 192 for more information You can re run any failed post processing job 172 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Match Scoring Chapter 6 Post Processing Tasks Order of Running Post Processing Administrative Tasks Run the post processing administrative tasks in this order 1 Match Scoring 501 Alert Creation 502 503 Update Alert Financial Data Alert Scoring 504 Assignment 505 Auto Close 506 Automatic Alert Suppression 507 Highlight Generation No ON OQ Ui A YW N Augment Trade Blotter j Score Trade Blotter Historical Data Copy 12 Alert Correlation 508 If the Oracle client has implemented the AA PTA application the Account Apptoval Pre Trade Approval Tasks may be run in parallel with any of the post processing administrative tasks 1 12 Before running the AA PTA application for the first time the AA PTA tasks must run in this order 1 Load Employee User Groups 2 Request Assignment Thereafter the frequency of when the AA PTA tasks are to be executed are determined by the Oracle client and m
259. efining the Auto Close Alert Alpotithm acese nies eere inpet ERR tere e ere Fe REAREN 178 To Set Up Anto Close RUS ue E EEEE EEE RU ERE ome Ea pde 179 To View ALLA ert Glosing ZAWrIDWIs a ue ei tecti ted ette eee ive Drev ted vete e Rod cot eee ctr 180 Sample Auto Closing Alert Rule 5 ror e tbt tabe nde te D test ote ust 181 Running the Auto Close Alett eet rtt tnter tot teen ek to eee io release tes verae osea 182 Io Rum Auto Close Mel soccer ei c i co DH e Fre Rr Orr EORR 182 Automatic Alert SUppressiOn s icscsssatasscrsssarsssenssasccesestssssensstsssssstsananssvensauatvasstvanssesccusbesssvensesenssvanssuansevensobessducuensedveveoes 183 Defining the Suppress Alert Alonte Ae tipici tee testen teer le petiere UE snes 183 Running the SuppressiOnJOb eee e e as tete Ea Pea o EAEE ENA quo e ete aono eddie pot 183 To RAW the Suppression 0b c eer ttn nette nie erm emat e rp er dtl i te 183 Highlight Generation Bn 184 Avioment Trade Blotter EM MM 184 Succus P MM M 185 Historical Data P NI PI bias Mesdvsadues Seadeua E E ehadss Neveu ch aa EAE AeA E a 185 Alert Cottelation ina 186 Running the Alert Correlation JOD visscisssissssveiersessssessacssaesseastocsedeteasiveadevedeaniadesonstasnsasessenspedsetesbensnsesienssensevetasess 186 Understanding Alert Correlation Configuration esssse
260. ege of accessing alerts that belong to the mapped scenario class Mapping of one or more Case Types Subtypes to a user or organization gives them the privilege of accessing cases that belong to the mapped Case Type Subtype Mapping of one or more correlation rules gives the privilege of viewing the correlations generated based on the mapped correlation Other parameters such as Line Organization Own Case Flag and Own Alert flag can be selected in the corresponding drop down list mentioned in the screen and can be updated by clicking the Save button Note The Own Alert and Case flag is required for taking ownership of the alerts and cases If an alert user needs to perform a Promote To Case action then the following pre requisites should be fulfilled 1 The user should be mapped to any one of the following user groups e Case Supervisor e Case Analyst e Case Analyst2 44 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide About Scenario Manager Login Accounts Chapter 3 Security Configuration 2 The user s Case Own flag should be enabled by setting the value to Y Or The user should be mapped to the Case Initiator Role About Scenario Manager Login Accounts Behavior Detection users gain access to the Scenario Manager application based on the following User ID and password authentication enables access to the Scenario Manager e An associated functional role corresponds to
261. egin Figure 51 Post Watch List Workflows Insurance The application completes these workflows in two sequences processing of a workflow requires data from a predecessor Refer to the note in the section Informatica Miscellaneous Post Watch List Workflows AML Brokerage on page 140 for more information Table 68 describes the Informatica Miscellaneous Post Watch List workflows in Figure 51 Table 68 Post Watch List Workflows Insurance Sequence Number Folder Workflow Number Workflow Name Predecessor 20 ORION Common 3501 w ph3501 Exp and Risk Review TPI ORION Common 2200 21 ORION Common 3502 w ph3502 Flag Trusted Trxn ORION Common 3501 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 169 Insurance Workflows Chapter 5 Informatica Workflows Informatica Summary Workflows Insurance Informatica Summary workflows maintain monthly aggregations of customer activity Figure 52 illustrates Informatica Summary workflows for Insurance Sequence 16 Sequence 17 Sequence 18 Note Processing can occur simultaneously within each sequence Sequence 19 Workflows with predecessors must wait for predecessor results before workflow processing can begin Figure 52 Summary Workflows Insurance The application completes these workflows in four sequences processing is not dependent on data from external predecessors The workfl
262. ement CaseSubClass1 KDD CASE SUBCLASS TAG CASE SUB CLASS LVL1 CD Element CaseSubClass2 KDD CASE SUBCLASS TAG CASE SUB CLASS LVL2 CD Element CaseTitle KDD REVIEW CASE CASE TITLE Element CaseDescription KDD REVIEW CASE CASE DESCRIPTION Element OwnerUserName KDD REVIEW OWNER OWNER ID Element ClosingAction KDD REVIEW CLS ACTVY TYPE CD Element CloserUserName KDD REVIEW OWNER OWNER ID Element DateEnteredStatus KDD REVIEW STATUS DT Element CreatorUserName KDD REVIEW OWNER OWNER ID Element OwnerUserName KDD REVIEW OWNER OWNER DSPLY NM Element Actions ActionName KDD ACTIVITY TYPE CD ACTVY TYPE SHORT NM DateActionTaken KDD ACTIVITY START DT UserNameTakingAction KDD REVIEW OWNER OWNER ID OwnerUserName KDD REVIEW OWNER OWNER ID CaseStatus KDD REVIEW STATUS STATUS CD Element Linked Alerts Element AlertID KDD REVIEW LINK END REVIEW ID Covers all actions taken on case not just the closing action Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 235 Calendar Manager Utility Chapter 7 Batch Processing Utilities Calendar Manager Utility Directory Structure Logs After loading holidays into the KDD_CAL_HOLIDAY table and weekly off days into the KDD CAL WKLY OFF table you can use the Calendar Manager Utility to update and manage system calendars Use the utility s Java and shell scripts to connect to the database and perform processing The
263. ement 1 Change the password in the database server for this user User Services 2 Use the Password Manager Utility to change the password KDD WEB Patch Database Patch Installer 1 Change the password in the database server for this user User 2 Use the Password Manager Utility to change the password Database Utility User Database 1 Change the password in the database server for this user 2 Use the Password Manager Utility to change the password Informatica Work Data Ingestion 1 Use the Informatica pmpasswd utility to generate an encrypted flow Operator Infor version of the password Do not use an encrypted version that matica Repository has single quotes in it user ID that executes 2 Find the svrpwd password property in the Behavior Detection INSTALL DIR ingestion manager scripts env sh workflows file and replace password with the desired encrypted password Refer to the Oracle Financial Services Behavior Detection Platform Installation Guide Release 6 1 3 for more information Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 47 About Changing Passwords for System Accounts Chapter 3 Security Configuration Table 18 System Account Passwords Continued System Account Active Pages User KDD ALTIO Subsystem Active Pages Instructions For the NetViz component Change the password for the properties hibernate connec tion password and
264. emoval Table 36 references the files that the system writes to the backup subdirectory by component Table 36 Backed Up Files by Component Component DataFiles FDT XDP Data Loader XDP FDT MDT The firm subdirectory within the data subdirectory contains the extract transform and load subdirectories that correspond directly to the workflow steps that Firm data moves through during Data Ingestion The following sections describe each subdirectory The extract subdirectory within the firm subdirectory contains checkpoint data and working files for each preprocessor during preprocessing Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 105 Data Ingestion Directory Structure Chapter 4 Data Ingestion transform Subdirectory load Subdirectory inbox Subdirectory tibspool Subdirectory tibspool in Subdirectory Each preprocessor also maintains checkpoint files that enable it to recover after a failure and without the loss of data integrity an FDT removes the files after it successfully preprocesses its data When finished each preprocessor moves its final preprocessed files to either the transform subdirectory for processing by FDT or to the load subdirectory for loading into the database The XpP file type identifies files that the preprocessor creates The transform subdirectory within the firm subdirectory contains the FDT s checkpoint data and workin
265. enario into the incorrect environment the Scenario Migration utility enables you to mame your source and target environments On exttact you can specify the environment name to which you plan to load the scenario If you attempt to load it to a different environment the system displays a watning prompt The Scenario Migration Utility produces two log files Figure 64 on page 277 load log and extract 1og These files reside in the following location INSTALL DIR database db tools logs Using the Scenario Migration Utility Configuring the Scenario Migration Utility This section covers the following topics which describe configuring and executing the Scenario Migration Utility including extracting and loading metadata e Configuring the Scenario Migration Utility e Extracting Scenario Metadata e Loading Scenatio Metadata The INSTALL DIR database db tools mantas cfg install cfg file contains common configuration information that Scenario Migration and other utilities require for processing Figure 64 provides sample information from the install cfg file that is specific to this utility Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 275 Scenario Migration Utility Chapter 8 Administrative Utilities HHHHHHHHHHHHHEHH SCENARIO MIGRATION CONFIGURATION H HHHHHHHHHHHHHHHHH HHHH GENERAL SCENARIO MIGRATION SETTINGS Specify the flags for whether scoring rules and wrappe
266. ent and resides in a date subdirectory of the logs directory in YYYYMMDD format TRUE Log UseDateLog Specifies whether to put default log file for a component in a date subdirectory Otherwise it is placed directly under the logs directory TRU E Log InitDir Specifies the location of the properties file for configuring the common logging module install cfg config DB Specifies properties related to database access DB Connection Driver Indicates the JDBC driver class name oracle jdbc driver Orac leDriver DB Connection Specifies the number of connections initially 1 InitialConnections to allocate in the connection pool DB Connection Indicates the maximum number of 10 MaximumConnections connections in the connection pool You should correlate this parameter to the number of configured threads for the component DB Connection Timeout Identifies the number of seconds to wait 10 before timing out on a database connection attempt Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 89 Data Ingestion Directory Structure Chapter 4 Data Ingestion Table 32 Datalngest xml File Configuration Parameters Continued Property Name Description Example DB Connection NumRetries Specifies the maximum number of times to 5 attempt to connect to a database before failing MARKET Specifies properties related t
267. entifies the directory to clean up The data backup System removes date subdirectories in YYYYMMDD format from this directory Directory 1 N DaysToKeep Specifies the number of days to keep for this 3 directory The system does not delete date subdirectories with the latest dates 100 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Ingestion Directory Structure Chapter 4 Data Ingestion Table 32 Datalngest xml File Configuration Parameters Continued Property Name Description Example DButility Specifies properties used to configure various utility processes Valid utility names are SecurityMarketDaily SecurityFirmDaily PortfolioManagerPosition AccountChangeLogSummary CustomerChangeLogSummary AccountToCustomerChangeLogSummary CorrespondentBankToPeerGroup lt UtilityName gt NumberofThread S lt UtilityName gt SequenceBatchs ize Specifies the number of worker threads that the give component uses when processing data Specifies the batch size when retrieving sequence IDs for records generated by given component 4 10000 Watch List Service Specifies properties used to configure the Scan Watch List Web Service Timeout value Specifies how many seconds a call to the Watch List Service made through the scanWatchList sh script will wait for the service request to finish This value should be set to the longest wait time
268. ents refer to scripts Subdirectory on page 82 for more information tibspool Contains the in and backup subdirectories to organize the raw Market data files that the system creates from raw data that it extracts from the Market data feed refer to bspool Subdirectory on page 106 for more information Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 81 Data Ingestion Directory Structure Chapter 4 Data Ingestion bin Subdirectory jars Subdirectory scripts Subdirectory The bin subdirectory within the ingestion manager directory contains the programs that interface with the Market data feed to capture market and business client data and to stream that data to the Market Data server A run script in the scripts subdirectory launches each program refer to scripts Subdirectory on page 82 for mote information The jars subdirectory within the ingestion_manager directory contains Java programs that Ingestion Manager uses A run script in the scripts subdirectory launches each program refer to scripts Subdirectory on page 82 for more information The scripts subdirectory within the ingestion manager directory contains the UNIX Bourne Shell scripts to run and stop runtime components Executing a run script runs a new instance of a component Executing a stop script terminates an active runtime component that is running in polling mode Each script returns a terminati
269. eous Pre Watch List Workflows Insurance Informatica Watch List Workflows Insurance e Informatica Miscellaneous Post Watch List Workflows lInsurance e Informatica Summary Workflows Insurance Each section provides a graphic that illustrates the workflow An accompanying table desctibes the process by sequence workflow number and name and internal or external predecessors if any Informatica Miscellaneous Pre Watch List Workflows Insurance Optional Miscellaneous workflows perform processing in support of workflows in multiple functional areas Figure 49 illustrates an Informatica Miscellaneous Pre Watch List workflow for Insurance External Predecessor Sequence 2 Note Processing can occur simultaneously within each sequence Sequence 3 Workflows with predecessors must wait for predecessor results before workflow processing can begin Figure 49 Pre Watch List Workflows lnsurance The application completes these workflows in two sequences processing of a workflow requires data from one predecessor These workflows may proceed simultaneously in all sequences as soon as processing of any predecessors has completed Table 66 describes the Informatica Miscellaneous Pre Watch List workflows for Insurance in Figure 49 Table 66 Pre Watch List workflows Insurance Sequence Number Folder Workflow Number Workflow Name Predecessor 2 ORION Common 2190 w ph2190 fotps instn p
270. er Group AMANALYST2GRP AM Analyst III AM Analyst III User Group AMANALYST3GRP AM Supervisor AM Supervisor User Group AMSUPVISRGRP AM Executive AM Executive User Group AMEXCUTIVEGRP AM Internal Auditor AM External Auditor AM Internal Auditor User Group AM External Auditor User Group AMINAUDITRGRP AMEXAUDITRGRP AM Data Miner AM mantas Administrator AM Data Miner User Group mantas Administrator User Group AMDATAMNRGRP AMMANADMNGR Table 9 describes the Case Management Roles and corresponding User Groups Table 9 Case Management Roles and User Groups Role Group Name User group Code Case Analyst1 Case Analyst1 UserGroup CMANALYST1UG Case Analyst2 Case Analyst2 UserGroup CMANALYST2UG Case Supervisor Case Supervisor UserGroup CMSUPERVISORUG Case Executive Case Executive UserGroup CMEXECUTIVEUG Case Internal Auditor Case External Auditor Case Internal Auditor UserGroup Case External Auditor CMINAUDITORUG CMEXAUDITORUG UserGroup Case Viewer Case Viewer UserGroup CMVIEWERUG Case Initiator Case Initiator User Group CMINITIATRUG Case Administrator Case Administrator User Group CMMANADMNUG 28 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Setting up a User Chapter 3 Security Configuration Table 10 describes the Account Approval and Pre Trade Approval Roles and corresponding User Groups Table 10 Account Approval
271. er Utility follow these steps 1 Verify that the Behavior Detection database is operational tnsping database instance name 2 Verify that the INSTALL DIR database db tools mantas cfg install cfg configuration file contains the correct source database connection information 3 Go to the directory where the shell script resides cd INSTALL DIR gt database db tools bin 4 Start the calendar manager shell script set mantas date sh YYYYMMDD whete YYYYMMDD is the date on which you want to base the calendar fot exam ple enter November 30 2006 as 20061130 The utility then verifies that the enteted date is valid and appears in the correct format If you do not enter a date or enter it incorrectly the utility terminates and logs a message that describes the error The error message displays on the console only if you have output to the console enabled in the INSTALL DIR database db tools mantas cfg categories cfg configuration file Refer to Configur ing Console Output on page 210 for more information As previously discussed the Calendar Manager Utility retrieves information that it needs for updating business calendars from the KDD CAL HOLIDAY and KDD CAL WKLY OFF database tables It then populates the KDD CAL table accordingly Refer to the Oracle Financial Services Behavior Detection Platform FSDM Reference Guide Volume 2 for more information That is for each calendar name found in the KDD CAL WKLY OFF
272. er tetro abeat E D E RES 135 AML Brokerage Workflows i ette teet rrr reti tese rH EE NE Tag eee aee Pest pee be e EE AN 135 Informatica Miscellaneous Optional Workflows AML Brokerage eee 136 Informatica Miscellaneous Pre Watch List Workflows AML Brokerage eee 137 Informatica Watch List Workflows AML Brokerage eese ten tete 138 Informatica Miscellaneous Post Watch List Workflows AML Brokerage eee 140 Informatica Summary Workflows AML Brokerage eeeeseeeseeeeeeeneteetet ertet tenerent 143 Informatica Balances and Positions Workflows AML Btrokerage eee 145 AMT Banking Informatica Worltlows n oie tremite tsi titia bietet 146 Informatica Miscellaneous Optional Workflows AML Banking eee 146 Informatica Miscellaneous Pre Watch List Workflows AML Banking eee 147 Informatica Watch List Workflows AML Banking eeeeeeeeeeeenenntnt tenente ten tenentes 148 Informatica Miscellaneous Post Watch List Workflows AML Banking eee 150 Informatica Summary Workflows AML Banking eese tentent tentent 151 Broker Compliance Informatica Workflows eese nte tenentes 153 Informatica Miscellaneous Optional Workflows Broker Compliance e 153 Informatica Miscellaneous Pre Watch List Workflows Broker Compliance
273. ers is common Oracle Financial Services recommends installation of the subsystem on a server that has sufficient direct connected RAID disk storage for both the product and ingested data When requiring high availability configure dual servers to access shared disk storage The shared disk supports high availability because data that the primary server writes to shared disk becomes available to the Backup server and its components during failure recovery Because the Data Ingestion subsystem can use substantial I O bandwidth and requires constant disk availability Oracle Financial Services discourages the use of NFS mounted disk storage The following sections describe the Data Ingestion directory structure Directory Structure Descriptions Table 27 lists important subdirectories that compose the INSTALL DIR ingestion manager directory structure 80 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Ingestion Directory Structure Chapter 4 Data Ingestion Table 27 Directory Structure Description Directory Name Description bin Contains the programs that interface with the Market data feed to capture Market data and to stream that data to the MDS refer to bin Subdirectory on page 82 for more information config Contains files used to configure the Data Ingestion components refer to config Subdirectory on page 85 for more information data backup Contains
274. ers orion mantas6 1 3 database db tools logs highlight generator log log category REPORT location users orion mantas6 1 3 database db tools logs report log log category DATA ANALYSIS TOOL location users orion mantas6 1 3 database db tools logs data analysis tool log Figure 56 Sample Log Routing Information The bolded text in the above example console implies that a specific utility displays logging information at the console in addition to recording the information in the approptiate log file In Figure 56 Alert Purge and Calendar Manager display relevant utility information in addition to logging it If an entry in the categories cfg file does not already include this information you must add it manually including the comma 210 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide About Annual Activities Chapter 7 Batch Processing Utilities About Annual Activities Oracle Financial Services requires that you perform certain calendar management tasks at least annually loading holidays and weekly off days from an Oracle Financial Services client This ensures that the application has the necessary information for populating its own business calendars This section covers the following topics e Loading Holidays Refer to Loading Holidays on page 211 for more information e Loading Non business Days Refer to Loading Non business Days on page 213 for more information
275. ers the following topics e Directory Structure Refer to Directory Structure on page 215 for more information e Logs Refer to Logs on page 216 for more information e Precautions Refer to Precautions on page 216 for more information e Using the Alert Purge Utility Refer to Using the Alert Purge Utility on page 216 for mote information e Sample Alert Purge Processes Refer to Sample Alert Purge Processes on page 221 for mote information Table 79 provides the directory structure for the Alert Purge Utility Table 79 Alert Purge Utility Directory Structure Directory Description bin Contains executable files including the run_alert_purge sh shell script lib Contains required class files in jar format mantas cfg Contains configuration files for example install cfg and cate gories cfg in which you can configure properties and logging attributes logs Keeps the INSTALL DIR gt database db tools logs Alert Purge log file that the utility generates during execution data Keeps sq1 files for execution Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 215 Alert Purge Utility Chapter 7 Batch Processing Utilities Logs As the Alert Purge Utility performs alert detection activities it generates a log that it enters in the lt INSTALL_DIR gt database db tools logs Alert Purge log file the logging process time stamps all entrie
276. es 1 The system populates temporary tables in the database some scenarios depend on these tables for performance reasons 2 A network creation process generates and characterizes networks filtering the links that the system evaluates in the construction of these networks 3 A match creation process creates matches based on detection of specific sequences of events in data that correspond to patterns or the occurrences of prespecified conditions in business data The process also records additional data that the analysis of each match may require Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide About Oracle Financial Services Operations Chapter 1 The Oracle Financial Services Behavior Detection Platform Post Processing During Post Processing the detection results are prepared for presentation to users This preparation is dependent upon the following processes 1 10 A match scoring process computes a ranking for scenario matches indicating a degree of risk associated with the detected event ot behavior Refer to section Match Scoring on page 173 for more information An alert creation process packages the scenario matches as units of work that is alerts potentially grouping similar matches together for disposition by end users Refer to section Update Alert Financial Data on page 176 for more information An alert financial data update process records additional data
277. es Chapter 4 Data Ingestion Table 39 Miscellaneous Java based Utilities Utility Name HouseholdProfileDerived Table Name HH SMRY MNTH Predecessor AccountProfile Not In Parallel ExternallnvestmentAccount ToAccount ACCT Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 115 Miscellaneous Utilities Chapter 4 Data Ingestion Portfolio Manager Utility The Portfolio Manager utility is used to populate the portfolio manager positions This utility reads tables Account and Account Position populated while executing preprocessors refer to Beginning Preprocessing and Loading on page 57 and Informatica ingestion refer to Processing Informatica Workflows and other Utilities on page 69 and creates records to populate the PORTFOLIO_MGR_POSN table Commands to Execute runUtility sh lt Utility Name gt runDL sh lt Utility Name gt While executing these commands replace lt Utility Name gt with PortfolioManagerPosition Example runUtility sh PortfolioManagerPosition runDL sh PortfolioManagerPosition Change Log Processing There are two ways for the ingestion manager to create Change Log records e Client provides Change Log DIS files e Ingestion manager to generate Change Log records by comparing current day reference data records to previous day reference data records There are two ways by which ingestion manager generate Change Log Recor
278. essors These wotkflows may proceed simultaneously in all sequences as soon as processing of any predecessors has completed Table 59 describes the Balances and Positions workflows in Figure 42 Table 59 Balances and Positions Workflows Broker Compliance Sequence Number Folder Workflow Number Workflow Name Predecessor 22 9007 AccountPositionDerived 23 9008 AccountBalanceDerived 9007 24 9009 HouseholdBalance 9008 Informatica Miscellaneous Post Watch List Workflows Broker Compliance Informatica Miscellaneous Post Watch List workflows facilitate the application of customer supplied measures of risk to corresponding entities transactions and instructions Figure 43 illustrates Informatica Miscellaneous Post Watch List workflows for Broker Compliance Sequence 13 2172 Figure 43 Post Watch List Workflows Broker Compliance The application completes this workflow in one sequence processing of the wotkflows does not require data from external predecessors Table 60 describes the Post Watch List workflows for Broker Compliance in Figure 43 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 157 Fraud Detection Informatica Workflows Chapter 5 Informatica Workflows Table 60 Post Watch List Workflows Broker Compliance Sequence Number Folder Workflow Number Workflow Name Predecessor 13 ORION Common 2172 w ph2172 update jurisdiction in ag
279. estigation Schema entente 40 Adding Entries through Excel Ublodd u spere een eoo eR REIR ARI 40 Creating Case Type and Class Map in Investigation Schema sse 40 Adding Entries directly in the Table using seribt eee teet tiennent tenente tenete tenen 40 Adding Entries through Excel Upload ntt ttt ese tiere tta e Peeters RES 40 Creating Organizations in the Database iisssscisscisseiosedssnsesssessieveioescociosedussdbictessisoastesteasisondovsdsobbebusoasinotdosedessbeve tees 41 Creating Organization in the Database through scripts esee eene tentent tenente tenete tnn 41 Creating Organization in the Database through Excel Upload e eene tenen 41 Mapping Users To Access Control Metadata esses tenete nte tenete tenete teen 42 OVGANIRALION C M 44 TUNISU ON srr etit e e qu EE e e E OE ERE DH hy SRA RoE 44 LDUDTIEDnr 44 M LIEG 44 S INEST ATP NE 44 COrTelaltoW Aube e PROPER O SO AU EEUU tions N EEEE EE E ERARE POOR 44 PA AAI ONAL Parameters TN tae 44 About Scenario Manager Login Accounts niente nte tenente par iea starik epaia eiei iae iira 45 Creating Scenario Manager Login Accounts tosta tente tonto tosta tokens tonto tenet e asa aa 45 Jo Create The Database Lonn Actoni ei pe aet teet teli eee o E fe ti e ER T ERST etes 45 To Set Up an Account and Function
280. et The operation ID must be unique within the database The numbers may be any number not already in the table OPRTN SET ID 100 SCORE 75 AND STATUS NW OPRTN SET ID 200 AGE 30 Assign an ID to each operation within the already divided operations OPRTN SET ID 100 SCORE 75 OPRTN SET ID 101 STATUS NW OPRTN SET ID 200 AGE 30 Assign the next operation set to chain the operations togethet Optionally assign or close an activity type code and or comment to the operation Insert the rows into the KDD AUTO CLOSE ALERT table Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 181 Auto Close Chapter 6 Post Processing Tasks Table 71 KDD AUTO CLOSE ALERT AGE 30 Table 71 resembles the entries into the KDD AUTO CLOSE ALERT table for the AGE gt 30 auto close alert CLS ACTIV OPRTN SE XPRSN ORDE ALERT AT OPRTR_C VALUE_T NEXT OPRTN ITY TYPE T ID R ID TR ID D x SET ID DMN CD CD CMMNT TX 200 1 113000005 gt 30 NULL MTS MTS 203 Close if age greater 7 than 30 Note The NEXT_OPRTN_SET_ID is NULL because this operation set contains only one operation Table 72 shows how to set it to the next operation s ID within the operation set Table 72 resembles entries into the KDD_AUTO CLOSE lt 75 SCORE and STATUS AL NW auto close alert ERT table for the
281. ete You can run intra day processing and add or omit detection runs at the end of non end of day ingestion batch runs These cycles of detection should only run BEX and some TC scenarios They detect only against that day s data and or data for open batches dependent on each scenario against which each batch is running The last intra day batch should be configured as the end of day batch You must run a final end of day batch that detects on all data loaded into the database for that day not only looking at the batch that was last loaded The system can display these alerts on the next day If you want to use either types of intra day ingestion you must set up intra day batches and one end of day batch If you do not the FDT processes more market data than necessary and runs for a long period Table 24 provides an example of setting up the KDD_PRCSNG_BATCH table Table 24 Processing Batch Table Set up ONE Intra Day batch 1 1 NNN TWO Intra Day batch 2 2 NNN NNN Intra Day batch N end of day 3 NNN 74 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Informatica Ingestion Parameters Chapter 4 Data Ingestion Informatica Ingestion Parameters Table 25 describes the default Informatica ingestion parameters that processing sets at installation These parameters reside in the PMrootDir ParamFiles file Table 25 Informatica Ingestion Parameters File Vari
282. etti tetti te tinte tente tn tenent 68 Populating Marketiand Business Data Tables aede tete vertere tette ee 69 Processing Informatica Workflows and other Ut lities eese eet tte tnnt tenente tenete 69 Intra Day InpestiOB PEOCESSITID ect ertetee tee oe eue orto ta EOR EERE etn tuerit dei de cse dried 73 Informatica Ingestion Parameterss sscissieseassvossasesscnsneasovessonseseioeseiedesaivandovedessanstionetosvsnsesadussiededesvobndosannstdehasesaseaedsiens 75 Alternatives to Standard Data Ingestion Practices eene ntes 77 Data Ingestion Directory SEEUCDUTO stie eerte dence dedic eiie tedesco dede e feat tei teed 80 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide vii Contents Directory Structure Descriptons issie tr teorici cei eise eene eere troie se react eee este eee varese 80 bin SUDAGIRECEOEY T 82 Jars Subditectoty co eee REI suena au RENE RERRRI RN UU EO ERE ERR TEE EI E REO LIES 82 SELIPtS SUDGIFECKORY d 82 conto Subdifectoty iicet ete ioter rp RITU Te oe Sc ED ue ido ITE VR HARDER 85 Data Ingest Custom XML Configuration I le sees eese esent tette tenete tette teen nente 86 Data Ingest Properties Configuration Fes tete iiti gi f ipie ae 87 Dita ML SER DD Nm 88 data SUDGIPECORY isnie i EEE EREE EEEE EE E E EE RE Me Fe ovs 103 dataj error SUONEN a a i a Mak hd E A ON om Oe 103 dall market Su bdire d
283. ewly generated alerts that match an active suppression rule The system can suppress alerts with the status of NEW based on their creation date generating scenario and focal entity The algorithm evaluates all candidate alerts and suppresses each alert that satisfies the criteria The suppressed alerts to which the system assigns a status of Closed remain for audit purposes and are still available for display for example through the Entity History page and processing for example reopening an alert Defining the Suppress Alert Algorithm The Suppress Alert algorithm does not suppress locked alerts The system locks an alerts while an analyst takes an action on it and then unlocks the alert when the analyst releases it The system skips all locked alerts until the next time it runs the Suppress Alert component When a user takes an action on an existing alert to suppress future alerts the suppression rule populates the KDD AUTO SUPPR ALERT table with the ctitetia for automatically suppressing and canceling suppression of the alerts e Refer to the Oracle Financial Services Behavior Detection Platform User Guide Release 6 1 3 for detailed information about initiating and canceling Alert Suppression e Refer to the Oracle Financial Services Behavior Detection Platform FSDM Reference Guide Volume 2 Release 6 1 3 for information about the KDD AUTO SUPPR ALERT table Running the Suppression Job To Run the Suppression Job The supp
284. f data and the nature of the behavior that the system is to detect The following sections describe each of the major steps in a typical production processing cycle e Start Batch e Data Ingestion e Behavior Detection e Post Processing e End Batch Using the Batch Control Utility you can manage the beginning of an Oracle Financial Services batch process Refer to Chapter 7 Batch Control Utility on page 223 for more information The Oracle Financial Services Ingestion Manager controls the data ingestion process The Oracle Financial Services Behavior Detection Platform Data Interface Specification DIS contains a definition of each solution set The Ingestion Manager supports files and messages for the ingestion of data Data ingestion involves receiving source data from an external data source in one of these forms The Ingestion Manager validates this data against the DIS applies required derivations and aggregations and populates the database with the results Refer to Chapter 4 Data Ingestion on page 51 for more information During Behavior Detection Oracle Financial Services Detection Algorithms control the scenario detection process The Detection Algorithms search for events and behaviors of interest in the ingested data Upon identification of an event or behavior of interest the algorithms record a match in the database The application executes the following processes in this order to find and record scenario match
285. f the dataset you customized was one of the items changed by the hotfix It is also possible that the hotfix changes the fields it expects from the dataset you customized causing the scenario to fail For scenarios you have customized you should always test the scenario hotfix without your customizations in place then re apply them to the scenario if necessary 312 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide APPENDIX C Report Management This appendix contains the configuration and integration details of the Report Pramework with Oracle Financial Services user interface The term client app is consistently used to Refer to the application that must be attached to the Report Pramework It provides online approval and archival capability Report Management The Report Management is an Active Pages interface that allows you to retrieve a list of reports based on the report criteria that are approved or awaiting approval You can select from the list of report to display either the live version of the scheduled reports awaiting approval or archived versions of reports that have been approved In addition you can view comments associated to a report Report Management enables to access a report based on the user s viewable organization jurisdiction and role Each report is mapped to one or more user roles set up during deployment Report Management is accessible as an application on the Active
286. f their life cycles occutred For end of day processing this may not be an issue For Intra day processing subsequent files should contain data in an ever increasing time sequence That is the first set of files should contain data from 09 00 00 to 11 00 00 the second set of files Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 111 Data Ingestion Archiving Chapter 4 Data Ingestion should contain data from 11 00 00 to 12 00 00 and so on This only affects events in a single order or trade s life cycle For example Batch 1 contains the following events e NW order event for order X timestamp 09 30 00 e MF order event for order X timestamp 09 45 00 Batch 2 contains the event NW order event for order Y timestamp 09 40 00 This order event is not considered out of sequence processing continues normally Rejection During the Loading Stage The last stage of ingestion is Loading At this stage the Ingestion Manager loads orders executions and trades into the database The Ingestion Manager rejects records during Loading if configuration of the database is incorrect for example setup of partitions are incorrect for the data being ingested Data Ingestion Archiving During ingestion processing the system moves processed files into an archive directory Firms can use these files to recover from processing malfunctions and they can copy these files to off line media for backup purposes
287. finition file The query definition file itself can be found here INSTALL DIR behavior detection algorithms share xml querydefs QBD BusEntityPaths xml The PATH ID and PATH NM in the table above are simply used to establish unique identifiers for this path The above paths may not necessarily apply to all types of alerts and they may have different levels of importance depending on what types of alerts they are applied to This variance is defined by a path configuration which is stored in the KDD BUS NTITY PATH CFG table Its structure is as follows Table 75 KbbD BUS NTITY PATH CFG Metadata Table Column Name Primary Key Foreign Key Column Type Nullable Y N Default PATH CFG ID NUMBER 10 No PATH ID KDD BUS NTITY PATH PATH NUMBER 10 No _ID SCNRO_ID KDD SCNRO SCNRO ID NUMBER 10 Yes SCNRO CLASS CD KDD SCNRO CLASS SCNRO C VARCHAR2 3 Yes LASS CD PRSDNC NB NUMBER 10 Yes Correlation Rules We can choose to apply the path identified by the PATH ID in this table to only alerts of a certain scenario or scenario class This is established by populating either the SCNRO ID or the SCNRO CLASS CD column respectively If neither of these columns are populated this path configuration is considered for an alert of any scenario or scenario class The importance or strength of a correlation determined by this path may vary depending on the scenario or scenario class of th
288. flow Number Workflow Name Description 3041 w ph3041 create addresses from ins This workflow maintains the addresses in the Derived t Address table It derives the addresses from the INSTRUCTION table 3120 w ph3120 create external entities This workflow maintains the External Entity table It from inst derives the entities from the INSTRUCTION table 3160 w ph3160 write inst associations t Loadthe Link Stage with any entity associations from ols instruction 3171 w ph3171 update account customer r This workflow updates the risk on the ACCT based on isk KYC Primary customer as well as other external risks It does not update the risk based on the EL process 3220 w ph3220 update bot activity risk This workflow updates the risk related values to all par ties in BOT 3230 w ph3230 update inst activity risk This workflow updates the risk related values to all par ties in INSTRUCTION MLM Brokerage Production Folder 0020 w p0020 dump csm to file This workflow can be used for the nightly dump of Cus tomer Summary Month Table to File current month 0030 w p0030 dump iasm to file This workflow can be used for the nightly dump of Insti tutional Account Summary Month Table to File Current Month 0090 0100 w p0090 reload csm from dump file w p0100 reload iasm from dump file This workflow can be used to rebuild Customer Sum mary Month Table from dump file This workflow can be
289. flows AML Brokerage cesses 145 Optional Workflows AML Banking eee 146 Independent Pre Watch List Workflows AML Banking 147 Watch List Workflows AML Banking eee 149 Post Watch List Wotkflows AML Banking eee 150 Summary Workflows AML Banking eene 152 Optional Workflows Broker Compliance sss 153 Workflows Broker Compliance sse 154 Summary Workflows Broker Compliance see 156 Balances and Positions Workflows Broker Compliance 157 Post Watch List Workflows Broker Compliance sss 158 Optional Workflows Fraud Detection sss 159 Pre Watch List Workflows Fraud Detection sss 160 Watch List Workflows Fraud Detection sse 162 Post Watch List Workflows Fraud Detection 164 Summary Workflows Fraud Detection eene 165 Pre Watch List workflows Insurance eee 166 Watch List Workflows Insurance eeeeeeeeenten tentent 168 Post Watch List Workflows Insurance eeeeeeeeeee 169 Summary Workflows Insuranree cceessesessssseseccesseesseensessseseneasecneaneees 170 xviii Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide List of Tables Table 70 Table 71
290. for AML Banking Predecessors 2005 3210 Misc Watch List se cmo idi Sequence 16 Sequence 17 0270 Sequence 18 Sequence 19 0480 0145 0470 0630 0640 0660 0750 Note Processing can occur within Sequence 20 0280 each sequence simultaneously 0146 ov 0760 Workflows with predecessors must wait for predecessor results before workflow processing can begin Figure 38 Summary Workflows AML Banking The application completes these workflows in five sequences processing of some workflows requires data from external predecessors refer to section Informatica Miscellaneous Optional Workflows AML Banking on page 146 and Informatica Watch List Workflows AML Banking on page 148 for more information These workflows may proceed simultaneously in all sequences as soon as processing of any predecessors has completed Table 55 describes the Informatica Summary workflows in Figure 38 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 151 AML Banking Informatica Workflows Chapter 5 Informatica Workflows Table 55 Summary Workflows AML Banking Sequence Workflow Number Folder Number Workflow Name Predecessor 16 MLM Banking Com 0081 w phO0081 truncate cbsm mon 16 ORION Production 0150 w p0150 truncate asms ORION Common 2005 16 9001 AccountDailyProfileTrade 16 9002 AccountDaily
291. fy that no jobs are active To clear the dispatch 1og file follow the steps 1 Shut down the dispatcher by following the procedure for Stopping the dispatcher Refer to section Stopping the Dispatcher on page 17 for more information Type cd KDD PRODUCT HOME 10gs at the system prompt where KDD PRODUCT HOME is your product server installation directory Type rm dispatch 1log to clear the dispatcher log Type start chkdisp sh sleep time and press Enter to restart the dispatcher To clear the job logs follow the steps 1 Stop the dispatcher by following the procedure for Stopping the dispatcher Refer to section Stopping the Dispatcher on page 17 for more information Type cd directory at the system prompt where directory is your log directory By default a job log resides in the directory KDD PRODUCT HOME 10gs You can configure the location of this log in the INSTALL DIR behavior detection algorithms MTS mantas cfg install cfg file If you do not know the location of the log directory check the install cfg file The log mantaslog location property indicates the log location the default is KDD PRODUCT HOME 10ggs but this location is configurable 3 Do either of the following Type rm job job id date time 1og at the log directory prompt to clear one job log where job id date time is the name of a specific job log Type rm job
292. g category GET DATASET QUERY location users orion mantas6 1 3 database db tools logs get dataset query log log category PUSH EMAIL location users orion mantas6 1 3 database db tools logs push email log ERATOR location users orion mantas6 1 3 database db tools logs highlight generator log log category REPORT location users orion mantas6 1 3 database db tools logs report log log category HIGHLIGHT GEN log category DATA ANALYSIS TOOL location users orion mantas6 1 3 database db tools logs data analysis tool log Specify the location of messages of a specific severity and category The valid values are the same as for category List multiple locations for each property by using a comma delimiter If an entry for a severity does not appear here the message is logged to the location specified for the category by the above property If that does not exist it is logged to the configured default location in install cfg NOTE The entry below is just an example It is commented out because mantas does not route by category severity These can be uncommented and modified if routing by category severity is necessary log EXAMPLE CATEGORY warning location syslog Figure 55 Sample Logging Information in the categories cfg File Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 209 About Administrative Utilities Chapter 7 Batch Processing Utilities Configuring Console Fi
293. g figures and tables provide also reflect this order Where predecessors exist processing of workflows cannot begin until completion of predecessor wotkflows In the figures of this chapter an artow between wotkflows indicates a processing dependency These dependencies may be internal to the workflow type for example Summary or Watch List or external to the workflow type for example Summary workflow dependent on Miscellaneous workflow Where applicable the figures and tables describe each type of dependency In Figure 29 for example processing can run workflow 2190 immediately after completion of workflow 2180 even if workflow 2070 has not completed Execution of the workflows according to the sequence number resolves dependencies 134 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide AML Brokerage Workflows Chapter 5 Informatica Workflows Workflow Tables Table 44 provides a list of columns and a description for each column provided in the Informatica workflow tables that each section provides Table 44 Informatica Workflow Table Descriptions Column Description Predecessor Indicator that processing of workflows cannot begin until com pletion of predecessor workflows Sequence Number Unique sequence number that indicates the order in which the workflows run Note Processing uniquely generates the sequence numbers that the figures represent Therefore they are
294. g files during transformation When finished the FDT moves its final transformed Firm data files to the load subdirectories for loading into the database The system writes the transformed data to files and then moves the files to the load subdirectory The FDT file type identifies the files that the FDT creates The FDT also maintains several checkpoint files that allow it to recover after a failure without the loss of data integrity The load subdirectory within the firm subdirectory contains additional subdirectories that contain preprocessed and transformed Firm data that the system queues for loading into the database Each loader component monitors its respective subdirectory that is data queue looking for data to load into the database a subdirectory exists for each kind of Oracle client data that processing loads into the database After loading data files into the database each loader moves the processed files to the backup subdirectory The inbox subdirectory within the ingestion_manager directory is an electronic mailbox or queue in which the Oracle client writes its data files for subsequent processing by Data Ingestion subsystem Data Preprocessor components Each Market or Firm Data Preprocessor retrieves the file it is assigned to process from the inbox subdirectory and then moves the file to the appropriate extract subdirectory for preprocessing The DIS describes the naming convention and content of each data files that an Ora
295. g the Data Retention Manager ii ssssiscssisssscdsssivesiessdssdusoiestuasseassotedboeaesedussotiesonsssselessbesnvosionsdessiesedate aa i taia 244 Configuring the Data Retention Manager o oses E regas ET ga CETE GE RAE GRGE REGERES 244 Execntmne M Data Retention Manager esie t tesa teas ots afia e CO EE e E iE 245 Crealing Partliotss s etd o RR RR Ree oon a de IR P de sea tea RP 247 UVEZInIat ing DOTRBOHSTS ote edet estere cette eerte utpat e eti qe tette etd riae tre te btts 247 M Masniaining Indexes teu nteger ete d ee 249 Utility Work illo REM n 249 KODEDR MAINT ORR UN DAR stete tsm a a loa canes oth A eaten tene 249 KOD DR JOB pae essa sa un an nT no oe re DM E D DE 250 FDU WO RUN VT ODE ssec fedes enter eta fluid te ance ce tule hene bent bein inte DAD DE 251 Database Staustics Manapetmefitzssc eee da e ir enaar RA Per d etba dto dea duo doe Ape RECHNER e e eA o enn 251 SOO m I REY 251 Using Database Statistics Management dese epp eie e dre SIE Taser dole ashes 251 Flag Duplicate Alerts Utility iet toten tto tenon tret aris ssir n rarius SEE SE Rt 253 Using the Flag Duphcate Alerts Utility treten tt ethernet ottiene obire beret eii da 253 Executing the Flag Duplicate Alerts Utility esent tnteentenenens 253 To Execute the Flag Duplicate Alerts Utility eee eee eee tentent tentent tentent tenete tenen 253 Notification Ee E ELE 254 Event Based iicet tetti
296. g ways e Adding Entries directly in the Table Using Script e Adding Entries through Excel Upload To add entries in the table using scripts follow these steps 1 Add the appropriate record to the KDD_CASE_TYPE_SUBTYPE database table 2 Add records to the table by using a SQL script similar to the following sample script insert into KDD CASE TYPE SUBTYPE CASE TYPE SUBTYPE CD CASE TYPE CD CASE TYPE NM CASE TYPE DESC CASE SUBTYPE CD CASE SUBTYPE NM CASE SUBTYPE DESC CASE CLASSIFICATION CD LAST UPDATED BY LAST UPDATED DT COMMENTS values AML SURV AML Anti Money Laundering Anti Money Laundering SURV AML Surveillance AML Surveillance AML null null null Refer to Creating Jurisdiction in tbe Database on page 35 for the steps to perform the Excel Upload of Case Subtype The Excel template to be used is KDD CASE TYPE SUBTYPE xls Creating Case Subclass1 in Investigation Schema Adding Entries through Excel Upload You can create a Case Subclass1 in the database in the following ways e Adding Entries directly in the Table using script e Add the appropriate record to the KDD CASE SUBCLASS1 database table Add records to the table by following SQL script similar to the sample script insert into KDD SUBCLASS1 CASE SUBCLASS1 CD CASE SUBCLASS1 NM CASE SUBCLASS1 DESC LAST UPDATED DT LAST UPDATED BY COMMENTS values BSA Bank Secrecy Act Bank Secrecy Act null null null Refer t
297. ggregate trade to file roduction 16 9001 AccountDailyProfileTrade 16 9002 AccountDailyProfileTransaction 17 9005 AccountProfile 9001 9002 ORION Production 0150 17 MLM Brokerage P 0290 w p0290 join instl acct activity to iasms MLM Brokerage Production 0170 roduction MLM Br okerage_Production 0210 MLM_Brokerag e_Production 0220 MLM_Brokerage_Production 0230 MLM_Brokerage_Production 0240 MLM_Brokerage_Production 0250 MLM_Brokerage_Production 0260 17 MLM Brokerage P 0300 w p0300 join instl acct activity to csms MLM Brokerage Production 0160 roduction MLM Br okerage_Production 0210 MLM_Brokerag e_Production 0220 MLM_Brokerage_Production 0230 MLM_Brokerage_Production 0240 MLM_Brokerage_Production 0250 MLM_Brokerage_Production 0260 18 ORION_Production 0310 w_p0310_update_asm_for_daily_activity 9005 ORION_Production 0190 18 MLM Brokerage P 0330 w p0330 update iasm for daily activity MLM Brokerage Production 0290 roduction 18 MLM Brokerage P 0340 w p0340 update csm for daily activity MLM Brokerage Production 0300 roduction 18 ORION Production 0650 w p0650 build IASD from ASMS MLM Brokerage Production 0290 19 ORION Production 0630 w p0630 build ATxSD from ASMS ORION Production 0310 19 ORION Production 0750 w ph0750 truncate APTxSM 19 MLM Brokerage P 0280 w p0280 create clog activity records ORION Production 0310 roduction 144 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide A
298. gnment Auto Approve Reject PTA Requests run apprvl assign requ est sh run apprvl auto apprv P PTA BA OWNER ASGNMT Auto assigns newly submitted Account Approval and Pre Trade Approval requests to a pool of users POOL When assigning the requests to a POOL the module fetches new unassigned requests and assigns each request to a pool based on the auto assignment rules defined in CR Default Assignee parameter via the Manage Installation Parameter screen Refer to the Oracle Financial Services Behavior Detection Platform Configuration Guide for details on how to configure the CR Default Assignee parameter P PTA AUTO APPRV REJECT rejct sh This procedure is used to Auto Approve and or Auto Reject newly submitted Pre Trade Approval requests To enable dis able the auto approve functionality refer to the CR Pre Trade Auto Approve parameter via the Manage Installation Parame ter screen To enable disable the auto reject functionality refer to the CR Pre Trade Auto Reject parameter via the Manage Installation Parameter screen Refer to the Oracle Financial Services Behavior Detection Platform Configuration Guide for details on how to configure the CR Default Assignee parameter Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 193 Account Approval Pre Trade Approval Tasks Chapter 6 Post Processing Tasks Before running the AA PTA application for the first time the AA
299. guration Understanding SMS Oracle Financial Servicesapplication SMS Engine is primarily responsible for user creation maintenance authentication and authorization As an administrator you can perform the following tasks Create users Manage users Create user groups Map user to user groups Assign roles to user groups Create functions Map functions to roles Accessing Oracle Financial Services Behavior Detection A user gains access to Behavior Detection based on the following Authentication of a unique user ID and password that enables access to Alert Management Case Management Account Approval Pre Trade Approval and Administration Tool For accessing Alert Management Set of policies that associate functional role with access to specific system functions One or more associated organizational affiliations that control the user s access to alerts Relationship to one or more scenario groups Access to one or mote jurisdictions Access to one ot mote business domains For accessing Case Management Set of policies that associate functional roles with access to specific system functions Access to one or mote case types subtypes One or more associated organizational affiliations that control the user s access to cases Access to one or mote jurisdictions Access to one ot mote business domains 26 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Setting up
300. guration parameters 124 configuring 123 executing 126 G Get Dataset Query Dataset ID 274 Threshold Set ID 274 Get Dataset Query with Thresholds Utility 261 274 Get Dataset Query with Thresholds Utility executing 274 Get Dataset Query with Thresholds Utility using 274 group AccountCustomerRole 58 AccountEmailAddress 58 AccountGroup 58 AccountPhone 58 AccountRealizedProfitAndLoss 58 H highlight generation 172 184 Historical Data Copy 172 185 Hotfix effect customization 311 hotfixes E mail 311 Informatica 311 Ingestion 311 Performance 311 Post Processing jobs 311 Scenarios 311 Secure FTP 311 User Interface 311 Informatica 52 ingestion parameters 75 summary generation 70 Watch List processing 72 workflow processing 69 Informatica Identity Resolution API Fuzzy Name Matcher Utility 126 Informatica workflow 133 AML Brokerage 166 Anti Money Laundering AML Banking 146 Anti Money Laundering AML Brokerage 135 balances and positions 145 156 Broker Compliance 153 BSM_ Production Folder 318 Fraud Detection 158 164 MLM Banking Common Folder 319 MLM Banking Production Folder 319 MLM Brokerage Common Folder 319 MLM Brokerage Production Folder 320 Optional Miscellaneous 153 158 166 optional miscellaneous 136 146 ORION Common Folder 321 ORION Production Folder 323 Post Watch List 150 157 330 Oracle Financial Services Behavior Detection Platform 6
301. gure 56 displays a section of the sample categories cfg file from Figure 55 Note Output the log routing information in bold text log category ALERT PURGE location console users orion mantas6 1 3 database db_ tools logs alert purge log log category BATCH CONTROL location users orion mantas6 1 3 database db tools logs batch control log log category CALENDAR MANAGER location console users orion mantas6 1 3 database db tools logs calendar manager 1log log category DATA RETENTION MANAGER location users orion mantas6 1 3 database db tools logs DRM Utility log log category TRUNCATE MANAGER location users orion mantas6 1 3 database db tools logs truncate manager log log category COMMON UTILITIES location users orion mantas6 1 3 database db tools logs common utilities log log category EXTRACT location users orion mantas6 1 3 database db tools logs extract log log category LOAD location users orion mantas6 1 3 database db tools logs load log log category REFRESH TEMP TABLE location users orion mantas6 1 3 database db tools logs refresh temp table log log category RUN STORED PROCEDURE location users orion mantas6 1 3 database db tools logs run stored procedure log log category GET DATASET QUERY location users orion mantas6 1 3 database db tools logs get dataset query log log category PUSH EMAIL location users orion mantas6 1 3 database db tools logs push email log log category HIGHLIGHT GENERATOR location us
302. gyAndCommodityMarketDaily MailHandlingInstruction EnergyAndCommodityMarketCenter UncoveredOptionsDailyTradeSumar X FrontOfficeTransactionRemittanceDo y cument AccountPositionProfitAndLoss RelatedFrontOffice Transaction Inform AccountldentifierChangeHistory ation LoanOrigination EnergyAndCommodityMarketTrading AccountFees Session AccessEvents EnergyAndCommodityReportedMarke EmployeeFirmTransferHistory tSale EmployeeToSecuritiesLicense 5 AccountRestriction LoanOriginationCondition BackOfficeTransaction LoanOriginationConditionType ChangeLog LoanOriginationDocumentPrintLog InvestmentAdvisor LoanOriginationFeeDetail SettlementInstruction LoanOriginationNote SystemLogonToSystemLogonType LoanOriginationToService Borrower OptionsViolation RegisteredRepresentativeTradeCom mission BackOfficeTransaction must be loaded after the AccountManagementStage utility has been executed see Miscellaneous Utilities Processing of data in Group1 requires no prerequisite information dependencies for preprocessing Groups that follow however rely on successful preprocessing of the previous group to satisfy any dependencies For example Ingestion Manager does not run Group 4 until processing of data in Group 3 completes successfully Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 59 Process Flow Chapter 4 Data Ingestion Process Flow Processing bases the dependencies that determine grouping on the
303. h constraints gt NULL field column name lt Constraint feature is optional May contain one or more constraints gt CONSTRAINT field column name operator operator value filter value gt lt NULL gt lt NULLS gt lt TABLE gt The following XML code sample is an example of the correct structure 266 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Analysis Tool Chapter 8 Administrative Utilities TABLE name ACCT gt lt NULLS gt NULL field ACCT TYPE1 CD NULL field RGSTN TYPE CD gt CONSTRAINT field DATA DUMP DT operator EQ value 15 NOV 2006 gt lt NULL gt lt NULLS gt lt TABLE name ACCT gt This code executes the following queries SELECT sum case when ACCT TYPE1 CD is null then 1 else 0 end as NULL CTO sum case when ACCT TYPE1 CD lt gt ltrim rtrim ACCT TYPE1 CD then 1 else 0 end as SPACE CTO sum case when RGSTN TYPE CD is null and DATA DUMP DT 15 NOV 2006 then 1 else 0 end as NULL CT1 sum case when RGSTN TYPE CD ltrim rtrim RGSTN TYPE CD and DATA DUMP DT 15 NOV 2006 then 1 else 0 end as SPACE CT1 FROM ACCT a Join Counts Analysis A join identifies the relationship between two tables by common fields Checking for join counts determines the referential integrity between two or more tables Determine join counts as follows e Simple join between two or more tables Refer to Simple
304. h file e For 64 bit operating systems the maximum value should not be greater than 3500 MB e For 32 bit operating systems the maximum value should not be greater than 1800 MB Minimum values vary by component the env sh file specifies these values The config subdirectory within the data ingest directory contains the application configuration files as Table 30 describes Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 85 Data Ingestion Directory Structure Chapter 4 Data Ingestion Data Ingest Custom XML Configuration File e DatalngestCustom xml refer to section Dara Ingest Custom XML Configuration File on page 86 for more information e Datalngest properties refer to section Data Ingest Properties Configuration File on page 87 for more information e Datalngest xml refer to section Data Ingest XML Configuration File on page 88 for mote information The DataIngest properties and DataIngest xml files contain settings for IP addresses port numbers file paths file extensions and other runtime settings including an application s performance tuning parameters Property files within the config subdirectory contain database user IDs and encrypted passwords The config datamaps subdirectory also contains XML data maps for parsing input data and mapping processed data to fields in files and in databases The XML data maps are preset and do not require any mod
305. he Oracle Financial Services Architecture Chapter 1 The Oracle Financial Services Behavior Detection Platform Each tier can contain all subsystems Subsystems in turn include one or more components that are divided into small installable units A solution set requires installation of the associated Oracle Financial Services components Tiers The Oracle Financial Services solution has two tiers Oracle Financial Services Behavior Detection Platform defines a foundation for building Oracle Financial Services solution sets It provides core data mining services frameworks and tools The application also includes interface packages that abstract non standard or proprietary commercial off the shelf COTS products Deployment of multiple Oracle Financial Services solution sets can occur on a single installation Each Oracle Financial Services solution set Anti Money Laundering and Fraud Detection extends the Oracle Financial Services framework Each adds domain specific content to provide the required services for addressing a specific business problem It includes reusable domain artifacts such as scenarios input data transformation code and profiling scripts A solution set also provides the required presentation packages and custom application objects for supporting user interface functionality specific to the business domain Subsystems The application is composed of the following four subsystems Data Ingestion Provides data prep
306. he QUERIES section of the analysis file SQL Report Lists all of the SQL run to produce the other sections of the report Error Report Displays any errors that occurred when any of the queries were performed Troubleshooting the Data Analysis Tool Table 105 lists common Data Analysis Tool errors and their solutions Table 105 Troubleshooting Data Analysis Tool Errors Error Message Cause Solution java io FileNotFoundEx The system cannot Verify the install cfg file indi ception path amp filename find the file specified cates the correct path java lang RuntimeExcep Tables lt table 1 gt In the analysis xml file tion Tables lt table 1 gt and table 2 gt are remove duplicate join contents in and already joined in this the lt JOIN gt element lt table 2 gt fashion Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 273 Get Dataset Query with Thresholds Utility Chapter 8 Administrative Utilities Get Dataset Query with Thresholds Utility Processing uses the Get Dataset Query with Thresholds Utility to store a dataset query in the Behavior Detection database with the threshold names and not with the threshold values When the Behavior Detection engine executes a scenario it substitutes the correct threshold values in the SQL query before submitting it to the database Tracking of the query that executes in the database occurs only thr
307. he correct user credentials ON oO A UC Verify that you can view alerts in the UI The display of alerts signifies that installation of the system is correct and works as designed Note The alerts that you can view depend on your user privileges 130 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Use of Control Data Chapter 4 Data Ingestion Resetting the Environment Once the testing is complete you can reset the environment to its clean state by purging the test alerts refer to section Art Purge Utility on page 215 for more on purging the alerts and deleting the control data threshold sets The delete thresholds sh shellscript helps you to delete control data thresholds from the environment The delete thresholds sh script e is present in the INSTALL DIR database db tools bin folder e takes SCNRO ID mandatory and TSHLD SET ID optional as parameters e deletes the threshold sets that are NOT in the default Behavior Detection Product range of sequence IDs 113000000 to 117000000 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 131 Use of Control Data Chapter 4 Data Ingestion 132 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide CHAPTER 5 Informatica Workflows This chapter describes the derivation and aggregation of data through workflows in Informatica after the Oracle Finanical Se
308. heet attachments Configuration Parameters Using Batch Export INSTALL DIR alert management alert mgmt WEB INF classes conf mantas cfg install cfg Exported data is stored in the absolute path investigation pdf export dir INSTALL DIR database db tools mantas cfg install cfg e Setthe maximum number of pdf export threads pdf archival maxthreads 3 e Number of alerts cases pet export web service call pdf archival service batchsize 5 e URL ofthe Alert Management service alertmanagement service url http alert management server server port mantas context services AlertManagementService For example if the alert management subsystem is running on a server prod1 mantas com and server port is 1001 then alertmanagement service url should be set to http prodi mantas com 1001 mantas services AlertManagementService To start a Batch Export follow these steps 1 Verify that the Alert Management UI is running 2 Verify that the INSTALL DIR database db tools mantas cfg install cfg configuration file contains the correct source database connection information and correct configuration information for this utility 3 Execute the script zun pdf archival utility sh located in the db tools bin directory 4 Upon completion of a successful export you will receive a confirmation message If the data was not successfully exported then you will receive an etror message with details of the error 23
309. hese alerts as possible duplicates and adds a system comment in the Action History section of the Alert Details page each alert cross references the other For example Possible duplicate of alert xxxxx Executing the Flag Duplicate Alerts Utility To Execute the Flag Duplicate Alerts Utility Use the following procedure to execute the Flag Duplicate Alerts Utility To execute the Flag Duplicate Alerts Utility run the following script after the Alert Creator Assigner and Auto Close processes jobs have completed lt INSTALL_DIR gt database db_tools bin flag duplicate alerts sh The system writes log information for this process to the following location lt INSTALL_DIR gt database db_tools logs run_stored_procedure log Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 253 Notification Chapter 7 Batch Processing Utilities Notification Notifications appear on the UI on the Home page and help alert users to items requiring their attention Notifications can be classified into two categories depending on the method of generation e Event Based e Batch Based Event Based These notifications are always associated with an event Following are the event based notifications e New Case Creation notification Whenever a user manually creates a new case a notification is generated to the owner of the case and 1f owner is a pool then notification is generated to all the user
310. hey operate within the Oracle Financial Services Behavior Detection Platform include the following Oracle Financial Services Installer Installs and configures the Oracle Financial Services application at a specific deployment site The Oracle Financial Services Installer also installs and upgrades any additional Oracle Financial Services solution sets and requires access to deployment specific configuration information for example machine names and port numbers System Administrator Configures maintains and adjusts the system and is usually an employee of a specific Oracle Financial Services customer The System Administrator maintains user accounts and roles monitors data ingestion and alert management archives data loads data feeds and performs post processing tasks In addition the System Administrator can reload cache However the scenario description is not visible to the System Administrator Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide xxi Scope of this Guide About This Guide Scope of this Guide This guide describes the physical and logical architecture of the Oracle Financial Services application It also provides instructions for installing and configuring the application its subsystem components and required third party software for operation The Oracle Financial Services application provides the foundation for all its products Advanced data mining algorithms and sop
311. histicated pattern recognition technologies power the application It provides an open and scalable infrastructure that supports rich end to end functionality across all Oracle Financial Services solution sets The application s extensible modular architecture enables a customer to deploy new solution sets readily as the need arises xxii Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide How this Guide is Organized About This Guide How this Guide is Organized The Oracle Financial Services Behavior Detection Platform Administration Guide includes the following chapters Chapter 7 The Oracle Financial Services Behavior Detection Platform provides a brief overview of the application and its components Chapter 2 Oracle Financial Services Jobs provides an overview of the Oracle Financial Services Job Protocol and procedures for performing various tasks that relate to starting stopping and recovering jobs Chapter 3 Security Configuration covers the required day to day operations and maintenance of users groups and organizational units Chapter 4 Data Ingestion describes the operation and process flow of Data Ingestion subsystem components Chapter 5 Informatica Workflows describes the derivation and aggregation of data through workflows in Informatica after the data ingestion process completes Chapter 6 Post Processing Tasks explains how to customize features that affect
312. ial Services Behavior Detection Platform 6 1 3 Administration Guide 237 Calendar Manager Utility Chapter 7 Batch Processing Utilities Executing the Calendar Manager Utility To Start the Utility Manually Updating the xpp car Table Warning When you have configured the system to calculate alert and case age in Business Days the calendar date of the current system date and the calendar date of the alert or case creation must be included in the calendar As such if you are running with a business date that is substantially behind the current system date you should set the lookForward parameter for the calendar manager sufficiently high to ensure that the system date is included on the calendar Additionally if you have alerts that are open for a very long period you should set the lookBack parameter sufficiently high to include the dates of your oldest open alerts If the business calendar does not cover either of these dates the processing reverts to calculating age in Calendar days The utility connects to the database employing the user that the utils database username property specifies in the install cfg file Typically you manage the Calendar Manager Utility as part of automated processing You can run the utility either inside a batch process that is after calling the start mantas batch sh script or outside a batch You can start the utility manually by using the following procedure To start the Calendar Manag
313. ica Workflows Appendix D Informatica Workflow Details Table 120 Informatica Workflow Description Continued Workflow Number Workflow Name Description 0260 w p0260 aggregate trade to file This workflow performs Daily Aggregation of Trades This populates the Institutional account deal activity file and customer deal activity file 0280 w p0280 create clog activity recor This workflow creates Change Log records that indicate ds a change in an accounts activity level as measured by the sum of deposits withdrawals and trades over a configurable time period months 0290 w p0290 join instl acct activity t This workflow joins institutional account and customer o iasms summary files and load them into institutional account summary month stage table 0300 w p0300 join instl acct activity t This workflow joins institutional account and customer o csms summary files and load them into customer summary month stage table 0330 w p0330 update iasm for daily acti This workflow performs Update of Institutional Account vity Summary Month Table from its corresponding staging table 0340 w p0340 update csm for daily activ This workflow updates the Customer Summary Month ity Table from its corresponding staging table 1030 w p1030 update cbps from deal This workflow counts the records in the Deal table which has an end date greater than or equal to the cur rent date by custome
314. ifications Table 30 Application Configuration Files File Name Description DataIngest properti Property file that contains settings that are configured at es installation These settings are of the most interest to an Oracle client regarding modification refer to Table 31 DataIngest xml XML configuration file that contains settings that normally remain as is refer to Table 32 DataIngestCustom xm XML configuration file that contains overridden settings from 1 DataIngest xml The following sections describe each of these configuration files The client can modify the DataIngest xml file to override default settings that the system provides However this file is subject to change in future releases Therefore upon installation of a newer version the client must reapply any modifications in the current DataIngest xml file to the newer DataIngest xml file To simplify this process the DataIngestCustom xml file is available for use This file holds all site specific changes to the DataIngest xml file The client can override any settings in DataIngest xml by placing the modifications in DataIngestCustom xml After installing a newer version the client can copy the older Data1ngestCustom xml file to DataIngestCustom xml in the new installation 86 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Ingestion Directory Structure Chapter 4 Data Ingestion
315. ilar to the following sample script insert into KDD TYPE CLASS MAP CASE TYPE CLASS SEQ ID CASE TYPE SUBTYPE CD CASE SUBCLASS1 CD CASE SUBCLASS2 CD values 1 AML SURV BSA CMIR insert into KDD TYPE CLASS MAP CASE TYPE CLASS SEQ ID CASE TYPE SUBTYPE CD CASE SUBCLASS1 CD CASE SUBCLASS2 CD values 2 AML SURV BSA FBAR Refer to Creating Jurisdiction in tbe Database on page 35 for the steps to perform the Excel Upload of Case Type and Class Map The Excel template to be used is KDD TYPE CLASS MAP xls Note All template Excel files for Excel upload will be available in the following location ftpshare STAGE Excelupload AMCMLookupFiles 40 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide About Configuring Access Control Metadata Chapter 3 Security Configuration Creating Organizations in the Database Creating Organization in the Database through scripts There can be two approaches to create an Organization in the database e Creating Organization in the Database through scripts e Creating Organization in the Database through Excel Upload Add entries directly to the KDD_oRG table using a script 1 Add the appropriate record to the KDD_ORG database table which Table 15 describes Note The KDD_oRG table is empty after system initialization and requires populating before the system can operate Table 15 KDD_ORG Table Attributes
316. ile The default behavior is not to reject records for these reasons Some fields are restricted to contain only one of the domain values that the DIS defines The Ingestion Manager rejects records that contain some other value For example Ingestion Manager rejects any Order record that contains an Account Type other than CR CI FP FB ER IA EE or any Special Handling Code other than that in the DIS Rejection During the Transformation Stage The second stage of ingestion is Transformation At this stage the Ingestion Manager derives the order and trade life cycles and other attributes that are necessary for trade related surveillance The Ingestion Manager rejects order records during Transformation for the following reasons e New and Cancel or Replace order events if the order identifier and placement date combination already exists order identifiers must be unique during a given day e New order events for child orders if the referenced parent order is itself a child order only one level of a parent child relationship is allowed The Ingestion Manager rejects trade execution records for New and Cancel or Replace trade execution events if the trade execution identifier and trade execution date combination already exists Trade execution identifiers must be unique during a given day Other problems can occur that do not cause rejection of records but cause handling of the records to be different e Lost Events e Out of Seque
317. inable cancellation codes FDT BufferSize value Allows an override to the BufferSize value property for FDT FDT LookForFutureEventTimes value FDT UsePrevailingSale value Specifies whether to use the prevailing FALSE Data Ingestion uses the following three parameters when calculating the expected print price for executions A reported market sale is comparable to an execution when its size is in the same tier T hirdTierMax Any size bigger than this value is considered part of the fourth tier FDT ExecutionSizeThresholds Specifies the maximum size for the first tier 1000 FirstTierMax FDT ExecutionSizeThresholds Specifies the maximum size for the second 5000 SecondTierMax tier FDT ExecutionSizeThresholds Specifies the maximum size for the third tier 10000 Data Ingestion uses the next five parameters when calculating the marketable time with reasonable size attributes for an order Processing divides orders into small medium and large based on their remaining unit quantities M ediumMinPercentAtBest order s remaining unit quantity that must be available at the best price for execution to be considered reasonable The minimum percentage value must be represented in its decimal equivalent for example 1 0 100 FDT OrderSizeMarketability Specifies the maximum size for an order to 1000 MaxSmallSize be considered small FDT OrderSizeMarketability Specifies the maxi
318. inancial Services Behavior Detection Platform 6 1 3 Administration Guide 19 Performing Job Tasks Chapter 2 Behavior Detection Jobs Restarting a Job number of jobs running do not overload the system In the event a job run dies unexpectedly that is not through a caught exception but rather a fatal signal you must manually verify whether any jobs are in the RUN state but do not have a mantas process still running which would mean that the job threw a signal You must update the status code to ERR to restart the job To start a new job without the dispatcher follow the steps 1 Create the new job and job description through an Oracle Financial Services Job Editor The application automatically assigns a unique ID to the job when it is created Associate the job to a Job Template Group using the KDD_JOB_TEMPLATE table Execute the start mantas sh script with the following parameters Start mantas sh template id sd DD MON YYYY ed DD MON YYYY nd where the optional job parameters sd and ed start date and end date respec tively are used to constrain the data that an algorithm job pulls back For example if these parameters are passed into an Alert Creator job the Alert Creator considers only matches for a grouping that has a creation date within the range that the parameters specify After a job runs successfully you can no longer copy edit or delete the job Restarting a job is necessary
319. ing shell script runScoreTradeBlotter sh Note This script is part of the Ingestion Manager subsystem and resides in the INSTALL DIR ingestion manager scripts directory Historical Data Copy Behavior Detection maintains records that ate directly involved with detected behaviors in a set of archive or ARC tables The Historical Data Copy HDC process identifies the records against which the current batch s scenario runs generated alerts and copies them to the ARC tables The run hdc ksh and upd kdd review fin sh must run upon completion of all detection and other alert post processing for example scoring and assignment but before the system ends the batch with the following shell script end mantas batch sh Note This script is part of the Database Tools that reside in the INSTALL DIR database db tools bin directory The run hdc ksh shell script manages the HDC process This process connects to the database as the user that the truncate database username property identifies in the INSTALL DIR database db tools mantas cfg install cfg file This property should identify the zzges user a user in the database with write access to tables in both the Market and Business schemas Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 185 Alert Correlation Chapter 6 Post Processing Tasks To improve performance you can adjust two configurable parameters in the INSTALL DIR data
320. ing tasks that you can perform when configuring the utility or running it manually e Configuring the Calendar Manager Utility e Executing the Calendar Manager Utility e Updating the KDD CAL Table The INSTALL DIR database db tools mantas cfg install cfg file contains common configuration information that Calendar Manager and other utilities require for processing Refer to Figure 54 The following sample section from the install cfg file provides configuration information specific to this utility including default numerical values in the utility s two required parameters HHEHHHHHHHHHHHHH CALENDAR MANAGER CONFIGURATION H H HHHHHHHHHHHH The look back and look forward days of the provided date These values are required to update the KDD CAL table The maximum look back or forward is 999 days calendar lookBack 365 calendar lookForward 10 calendar lookBack Determines how many days to iterate backward from the provided date during a calendar update calendar lookForward Determines how many days to iterate forward from the provided date during a calendar update The maximum value that you can specify for either of these parameters is 999 days Note The lookback period should be at least 90 days and as long as any alerts are likely to be open The lookforward period does not need to be more than 10 days This is used when calculating projected settlement dates during Data Ingestion Oracle Financ
321. ion File 306 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Logging Configuration File Appendix A Logging Logging Location Property Values Log File Sizes The log category lt CATEGORY_NAME gt location property enables you to specify the location of a message for a specific category If you do not specify a location value the system logs messages in a default location Table 117 identifies the valid values for this property Table 117 Logging Location Property Values Property value Log location console Records the logs to the system out or system err file syslog Records the logs to a remote UNIX syslog daemon This is the default location eventviewer Records the logs to the Event Log system mantaslog Indicates the format of the mantaslog filename as job lt job gt datetimestamp if running the algorithms For other subsystems the format is mantaslog datetimestamp The file resides at the location that the log mantaslog location property specifies in the appropriate install cfg file If this property is unspecified the system outputs logs to the console lt path gt lt file Records the logs to a file with the filename lt filename gt which name gt resides at lt path gt For example log message library user jsmith message mes sages dat lt name address gt Records the logs in a message to the e mail address indicated by lt name addre
322. ion key must exist in the KDD JRSDCN table Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Setting up a User Chapter 3 Security Configuration Obtaining Information Before Configuring Access Control Before you perform access control activities for example adding a group modifying user information or deleting a user contact your system administrator for the following information to add to the locations in Table 12 Table 12 Access Control Items and Locations Data Item Location User Name KDD REVIEW OWNER User ID KDD REVIEW OWNER in the database Primary Organization KDD REVIEW OWNER Viewable Organizations KDD REVIEW OWNER ORG Role CSSMS ROLE MAST User to Scenario Group KDD SCNRO GRP ACCESS Scenario Group KDD SCNRO GRP ACCESS Scenario to Scenario Group KDD SCNRO GRP MEMBERSHIP Case Type Subtype KDD CASE TYPE SUBTYPE Business Domain KDD BUS DMN Jurisdiction KDD JRSDCN Email Address KDD REVIEW OWNER Note Email ID is mandatory for users who would need to take Email action The user ID should configured with valid email IDs while configuring the same through the User Maintenance UI Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 33 About Configuring Access Control Metadata Chapter 3 Security Configuration About Configuring Access Control Metadata
323. ions gt lt Lookback number 1 unit D gt lt Scenarios gt lt Scenario id 234 gt lt Scenario id 345 gt lt Scenarios gt lt ExistingCorrelationParams gt lt ExtendFlag gt TRUE lt ExtendFlag gt lt NonExtendableCaseStatuses gt lt CaseStatus gt CCL lt CaseStatus gt lt CaseStatus gt NVST lt CaseStatus gt lt NonExtendableCaseStatuses gt lt ExistingCorrelationParams gt lt Actions gt Scoring strategy MAX incStrategy ALERT COUNT gt lt CasePromotion gt lt FocusTypePref gt CU AC lt FocusTypePref gt lt AlertCorrAttrOperations gt CDATA CORR BUS NTITY ID 5 AND CORR PRECEDENCE NB lt 6 gt lt AlertCorrAttrOperations gt ExistingCasePromoteLossRcvryData TRUE ExistingCasePromoteLossRcvryData Case type AML subtype SURV subClassTagLevell CHK FRD subClassTagLevel2 ALTD INST lt CasePromotion gt lt Actions gt CorrelationRule e MinAlertCount required The minimum number of alerts involved in a correlation for it to be considered a valid correlation The minimum acceptable value is 2 e PrecedenceThreshold required Number indicating the maximum precedence value that a business entity shared between alerts must have in order to be considered a correlation by this rule The lower the precedence number the stronger the relationship Alerts will not be considered for the correlation unless the precedence number associated with the business entity to alert is
324. iorities are enabled in a hierarchical fashion i e if DIAGNOSTIC priority is enabled NOTICE WARN and FATAL are also enabled but TRACE is not Uncomment the desired log level to turn on appropriate levels Note DIAGNOSTIC logging is used to log database statements and slows down performance Only turn on if you need to see the SQL statements being executed TRACE logging is used for debugging during development Also only turn on TRACE if needed log fatal true log warning true log notice true log diagnostic true log trace true Specify whether logging for a particular level should be performed synchronously or asynchronously log fatal synchronous false log warning synchronous false log notice synchronous false log diagnostic synchronous false log trace synchronous true Specify the full path and filename of the message library log message library Specify where messages of a specific category should be logged The property name should be of the form log category CATEGORY _NAME location If logging to a category that is not specified below the messages are logged to a configurable default location The valid values are console syslog eventviewer mantaslog an e mail address or the full path to a file If mantaslog is specified the property log mantaslog location must be Gb H db H dk Specified with the desired path to the logfile If running the algorithms Con
325. ios It then groups them based on one of four implemented algorithms and a specified list of bindings for grouping It requires parameter settings to designate e Choice of grouping algorithm to use e Scenario types associated with the set of matches to consider for grouping e Bindings on which to base break group compatibility 174 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Alert Creation Chapter 6 Post Processing Tasks Grouping Algorithms When grouping algorithms choose from the following e BIND MATCH The Alert Creation module creates alerts based on matches with matching bindings values based on a provided list of bindings to use when determining groupability e BIND_BEHAVIOR_SCENARIO_CLASS The Alert Creation module creates alerts based on matches with matching scenario class code and with matching bindings values based on a provided list of bindings to use when determining groupability e BIND_BEHAVIOR_SCENARIO The Alert Creation module creates alerts based on matches with matching scenario ID and with matching bindings values based on a provided list of bindings to use when determining aroupability e BIND BEHAVIOR PATTERN The Alert Creation module creates alerts based on matches with matching pattern ID and with matching bindings values based on a provided list of bindings to use when determining aroupability e SINGLE ALERT MATCH The Alert Creation module creates
326. ire the use of the Four week Average Daily Share Quantity 4 wk ADT to define an order as small medium or large based on how it compares to a percentage of the 4 wk ADTV The 4 wk ADTV is derived on a daily basis by the process market summary sh script in the end of day batch once the Daily Market Profile is collected for each secutity from the relevant market data source For firms using the MiFID scenarios and running a single end of day batch the process market summary sh script must be executed prior to running the runFDT sh script such that the 4 wk ADTV for the Current Business Day incorporates the published Current Day Traded Volume Figure 13 depicts dependency between the process market summary sh script and the runFDT sh sctipt runFDT sh process market summary sh Figure 13 Dependency between process market summary shand runFDT sh For intra day batch ingestion or intra day execution of the MiFID scenarios the process flow does not change from Figure 12 Since the current day s 4 wk AD TV is not available until the end of the day the previous day s 4 wk ADT is used to determine order size For additional information on configuring the percentage values used to define a MiFID eligible order as Small Medium or Large refer to the Market Supplemental Guidance section in the Dara Interface Specification Release 6 1 3 56 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Process Flow
327. istration Guide Data Retention Manager Chapter 7 Batch Processing Utilities Daily Partitions Monthly Partitions 2 Weekly Partitions T Drop oldest partition s a 2 Create new partition s for the month Data Retention Manager uses special processing on the following tables to maintain open orders and associated executions and trades EXECUTION ORDR ORDR_EVENT ORDR_STATE_CHANGE TRADE TRADE_EXECUTION_EVENT O Drop oldest partition s 2 Create new partition s for the month At completion of processing for mixed date tables G Copy data for open orders from oldest the tables contain the following information partition s to temp table if applicable Current day and drop oldest partitions s Previous business day 4 Copy saved data to now oldest partition oe day of the previous two End of week partition s between the current Figure 63 Database Partitioning Process Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 243 Data Retention Manager Chapter 7 Batch Processing Utilities Using the Data Retention Manager The Data Retention Manager typically runs as part of automated processing that a job scheduling tool such as Maestro or Unicenter AutoSys controls However you can run Data Retention Manager manually on a daily weekly or monthly basis to manage database tables The following sections describe configuration and execution of the
328. ith daily net worth balance data 0550 w ph0550 aggregate IPSS from IT This workflow aggregates the Insurance transaction records into Insurance Transaction stage 0551 w ph0551 IPSS from IPB This workflow aggregates the Insurance policy balance into Insurance Transaction stage 0552 w ph0552 build IPSD from IPSS This workflow performs updates of Insurance Policy Summary Daily Table from the Insurance Policy Sum mary Month Stage table 0553 w ph0553 build IPSM from IPSS This workflow performs updates of Insurance Policy Summary Month Table from its corresponding staging table 0554 w ph0554 truncate IPSS This workflow truncates the Insurance Policy Summary Stage table 0630 w p0630 build ATxSD from ASMS This workflow maintains Account transaction summary daily from account summary stage table 0640 w p0640 build ATdSD from ASMS This workflow maintains Account trade summary daily from account summary stage table 0650 w p0650 build IASD from ASMS This workflow unlike its stated name builds institutional account summary daily from the corresponding staging table 0660 w p0660 FOTPSR to AASD This workflow populates ACCT ATM SMRY table from FO TRXN PARTY STAGE RISK table based on TRXN EXCTN DT 0750 w ph0750 truncate APTxSM This workflow truncates the ACCT PEER TRXN SMRY MNTH table 0760 w ph0760 ASM to APTxSM This workflow populates the ACCT PEER TRXN SMRY MNTH table from ACCT SMRY MNTH table 324 Oracle Financial Ser
329. ities Chapter 4 Data Ingestion Change Log Summary Processing ChangeLog This data loader is run after all preprocessors that produce ChangeLog records are complete and any ChangeLog DIS files are preprocessed i e runDP sh ChangeLog There are three different parts of ChangeLog Summary scripts that must be run They ate AccountChangeLogSummary run the files in the following order W runUtility sh AccountChangeLogSummary W runDL sh AccountChangeLogSummary CustomerChangeLogSummary run the files in the following order W runUtility sh CustomerChangeLogSummary W runDL sh CustomerChangeLogSummary AccounttoCustomerChangeLogSummary run the files in the following order W runUtility sh AccountToCustomerChangeLogSummary W runDL sh AccountToCustomerChangeLogSummary All of these scripts must be run after the ChangeLog data loader is run refer to Change Log Processing on page 116 for more information The AccountChangeLogSummary loads the ACCT cHG LOG SMRY table the CustomerChangeLogSummary loads the CUST CHG LOG SMRY table and the AccounttoCustomerChangeLogSummary loads the cUST ACCT CHG LOG SMRY table Network User Account Map Trade Blotter The Network User Account Map utility is used to populate the NTWK USER ACCT MAP table This table maps an employee to accounts in which the employee has a role It depends on the Employee Customer Account To Customer Account Customer Role and Account
330. ity dat analysis input Path and name for the XML input file By default this is the analy sis xml file under the install dir data base db tools mantas cfg directory You can override this at the command line dat analy sis input opt man tas database db tools mantas cfg analysis xml dat analysis output dat output format Path and file name of output file for the analysis report You can override this at the com mand line Output format for the report Acceptable output formats are HTML or TEXT dat analysis output opt mantas database db tools data analy sis html dat output format HTML dat output delimiter Not currently used The delim iter for the format TEXT is always a comma For additional information about the install cfg file Refer to install ofg File on page 198 for more information The analysis xml configuration file specifies the queries that you can use to analyze the data that the database schema provides You can perform the following types of queries Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 263 Data Analysis Tool Chapter 8 Administrative Utilities Analysis Constraints e Distinct Values for Fields of Interest Analysis Refer to Distinct Values for Fields of Interest Analysis on page 265 for more information e Null and Padded Space Count Analysis Refer to
331. ity generates during execution Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 241 Data Retention Manager Chapter 7 Batch Processing Utilities Logs Oracle stored procedures implement Data Retention Manager and conducts some logging on the database server A configuration parameter in the install cfg file controls the path to which you store the logs on the database server As the Data Retention Manager performs partitioning and indexing activities it generates a log that it enters in the lt INSTALL_DIR gt database db tools logs DRM Utility log file the logging process time stamps all entries The log file contains relevant information such as status of the various processes results and error records You can modify the current logging configuration for Data Retention Manager in the configuration files lt INSTALL_DIR gt database db_tools mantas_cfg install cfg and categories cfg For more information about logging in these configuration files Refer to Common Resources for Administrative Utilities on page 198 and Appendix A Logging on page 299 for more information Processing Flow Figure 63 illustrates the Data Retention Manager s process flow for daily weekly and monthly partitioning Based on a table s retention period the utility drops the oldest partition and then adds a new partition 242 Oracle Financial Services Behavior Detection Platform 6 1 3 Admin
332. iven utility must be executed after the Predecessor utility A utility listed in the Not In Parallel column means that the given utility can be executed before or after the Not In Parallel utility but it cannot be executed in parallel Table 39 Table 39 Miscellaneous Java based Utilities Utility Name Table Name Predecessor Not In Parallel AccountGroupProductAlloc ation ACCT GRP PRDCT ALLOC ATION AccountManagementStage ACCT MGMT STAGE AccountPositionDerived ACCT POSN AccountProfile InterestedPartyToEmployee INTERESTED PARTY EMP RegOToBorrower REG O BORROWER UncoveredOptionExposure Daily UNCVRD OPTNS EXPOSUR E DLY AccountBalanceDerived RegisteredRepresentativeP rofile RGSTD REP SMRY MNTH AccountDailySecurityProfile ACCT SCRTY SMRY DAIL Y AccountDailyProfileTransac tion ACCT TRXN SMRY DAILY AccountBalanceDerived ACCT BAL POSN SMRY AccountPositionDerived AccountProfile AccountDailyProfileTrade ACCT TRADE SMRY DAIL Y AccountDailySecurityProfile HouseholdBalance HH BAL POSN SMRY AccountBalanceDerived InvestmentAdvisorProfile NVSMT MGR SMRY MNTH AccountManagementStage AccountProfile ACCT SMRY MNTH AccountDailyProfileTrade AccountDailyProfileTransac tion 114 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Miscellaneous Utiliti
333. kflow is used for the daily aggregation of the le day trades data This populates the account summary file 0380 w p0380 aggregate bot inter hh jrn This workflow is used for the daily aggregation of the ls to file back office transaction data This populates the man aged account summary file 0390 w p0390 aggregate trades to rrst This workflow aggregates trades into registered repre sentative security 0400 w p0400 aggregate nm inter hh jrnl This workflow is used for the daily aggregation of the S to file back office transaction data It calculates Investment Manager Summary Journal Data 0410 w p0410 aggregate nm net worth to This workflow is used for the calculation of the Daily file change in Delinked Sub Accounts and Net worth It loads Investment Manager Summary Journal Data 0420 w p0420 aggregate nm third party t This workflow is used for the daily aggregation of the rxns to file front office transaction data It calculates the Investment Manager Summary Third Party Data elements 0430 w p0430 join daily activity to mas This workflow joins three flat files that contain the data ms for Inter HH Journal and block allocation trades together The result is inserted into the MANGD ACCT SMRY TEMP table 318 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Informatica Workflows Appendix D Informatica Workflow Details Table 120 Informatica Workflow Description Continued
334. l comma sepatated list of current alert status codes The utility then purges actions then matches then alerts according to the contents of the KDD AP ACTION KDD AP MATCH and KDD AP ALERT tables The utility captures purging results and any errors in the Alert Purge 1og file Note The Alert Purge Utility does not purge any data from archive tables for erroneous alerts Also the system does not update score and previous match count values associated with generated matches and alerts since creation of the erroneous matches The Alert Purge Utility has an automatic restart capability in that any interruption in the purge processing resumes at that point regardless of the input parameters The system documents logs information about the interruption in the INSTALL DIR database db tools logs Alert Purge 1og file Otherwise any restart that has not progressed to the purge component behaves as a new processing run The restart capability allows interrupted purges to resume at a convenient point but is unable to execute all desired input parameters 220 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Alert Purge Utility Chapter 7 Batch Processing Utilities Sample Alert Purge Processes Table 81 Example of Matches and Alerts Associated with Purge Alerts This section includes three examples of the Purge Alerts process based on input parameters In these examples the process executes
335. le fotps It derives the records from the FOTPS table MLM Banking Production Folder 0080 w p0080 dump cbsm to file This workflow can be used for the nightly dump of Cor respondent Bank Summary Month Table to File current month 0140 w p0140 reload cbsm from dump file This workflow can be used to rebuild Correspondent Bank Summary Month table from dump file 0145 w ph0145 truncate CBPTxSM This workflow truncates the CB PEER TRXN SMRY MNTH table 0146 w ph0146 CBSM to CBPTxSM This workflow populates the CB PEER TRXN SMRY MNTH table from CLIENT BANK SMRY MNTH table MLM Brokerage Common Folder 2071 w ph2071 truncate das This workflow is used to Truncate the DAILY AGG STAGE table 2122 w ph2122 aggregate tsv offsetting This workflow populates DAILY AGG STAGE table with trades to tdtcs aggregated TRADE Data DAILY AGG STAGE table in turn is used to populate OFFSETING ACCT PAIRS and TRADE DAILY TOT CT STAGE tables 2192 w ph2192 update inst instn seq id This workflow marks all institutions with an Oracle Financial Services generated INTSN SEQ ID in INSTRUCTION Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 319 Informatica Workflows Appendix D Informatica Workflow Details Table 120 Informatica Workflow Description Continued Work
336. le except account 3191 w ph3191 apply risk to acct entiti This workflow applies all risk related values to ACCT es table 3200 w ph3200 apply membership to entit This workflow maintains the membership tables based ies on the current WL processing results 3210 w ph3210 update fotps activity ris This workflow updates the risk related values to all par k ties in FOTPS 3240 w ph3240 update INS activity risk This workflow updates the risk related values to all par ties in INSURANCE 3501 w ph3501 Exp and Risk Review TP This mapping sets the status of a Trusted Pair to expire based on its Expiry Date Also if S TP RISK REVIEW FLAG is set to Y then this map ping reviews updates the risks for IA and EE parties associated with trusted pairs to reflect the latest risk as in the base tables If they have increased by substantial amount to move them to a next risk zone it is recom mending risk cancellation RRC 3502 w ph3502 Flag Trusted Trxn This mapping flags the transactions as trusted or not trusted based upon entry in the k d trusted pair and kdd trusted pair mbr tables It only looks at today s transactions ORION Production Folder 0010 w p0010 dump asm to file This workflow is used for the nightly dump of Account Summary Month Table to file for the current month 0070 w p0070 reload asm from dump file This workflow can be used to rebuild ACCT Summary Month table from the dump file This wo
337. les the job runs After a job runs successfully you can no longer copy edit or delete the job The start mantas sh script waits for all jobs in the template group to complete Note Before you run Network jobs change the batch size of Network related counter to 10 000 By default the counters is set to 100 The network counters in KDD COUNTER are NTWRK ID SEQ NODE ID SEQ LINK ID SEQ LINK SUMMARY ID SEQ and LINK TYPE SUMMARY ID SEQ Refer to the Oracle Financial Services Behavior Detection Platform FSDM Reference Guide Vol 2 Oracle Financial Services Data for more details Starting Jobs without the Dispatcher Clients who use multiple services to run jobs for one database must run the jobs without dispatcher processes If the client does use dispatchers on each machine each dispatcher may run each job which causes duplicate detection results To run a job template without a dispatcher add the parameter nd to the command line after the template ID For example start_mantas sh 100 nd This causes the start_mantas sh script to execute all jobs in the template rather than depending on the dispatcher to run them The jobs in the template group run in parallel The dispatcher can ensure that it is only running a set number of max jobs at any given time so if the max is set to 10 and a template has 20 jobs associated to it only 10 run simultaneously When running without the dispatcher you must ensure that the Oracle F
338. less than or equal to lt the value defined e AlertAttrOperations optional Defines operations used to further constrain the alerts to be used for correlation An operation consists of an alert attribute identified by ATTR NM compared to a string literal for example a from alert and 7o alert can be correlated if they both have JRSDCN CD AMEA represented by BOTH JRSDCN IN AMEA IND above or an alert attribute compared to the same attribute for example a from alert and to alert Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 189 Alert Correlation Chapter 6 Post Processing Tasks can be correlated if FROM SCORE_CT TO SCORE_CT The set of supported comparison operators are lt gt lt gt IN and NOT IN Note that because the SCNRO_ID attribute of both alerts and correlations can potentially have multiple values only the IN and NOT IN operators should be used in expressions involving SCNRO_ID The rest of the operators can only support a single value operands Also there should be no space in the scenario id list specified For example BOTH SCNRO_ID IN 115600002 114690101 is correct and BOTH SCNRO_ID IN 115600002 114690101 is incorrect Multiple operations can be strung together by logical AND and OR operators and operation precedence can be defined with parentheses Note that the text of an AlertAttrOperation must be wrapped in a CDATA tag as above
339. lid column name from the MANTAS schema tables KDD REVIEW and KDD REVIEW FINANCIAL e The CDATA section of the XML tag AlertCorrAttrOperations is validated as follows m The Correlation Alert operation should be CORR m The valid operators can be gt lt gt IN NOT IN AND and OR Note The SCNRO ALERT CT attribute works fine when used with the IN or NOT IN operators The alert correlation job gives an error when the SCNRO ALERT CT attribute is used with operators like gt gt lt lt andl This attribute is unlikely to be used in a correlation expressions but if it is used then it is recommended to use only with the IN or NOT IN operatots m The expression can be nested to any arbitrary length provided it confirms to a general syntax Operand Operator Operand Logical Operator Operand Operator Operand For Example a CORR SCNRO ID gt 0 b CORR SCNRO ID 0 AND CORR SCNRO ID CORR SCNRO ID Note space character is expected between each Operand and Operator combination Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 293 Alert Correlation Rule Migration Utility Chapter 8 Administrative Utilities m The precedence of an operation may be depicted using a pair of parenthesis iC and 2 m The Correlation Alert attributes provided should be a valid column name from the MANTAS schema tables KDD ALERT CORR and KDD ALERT CORR SCNRO Loading Ale
340. lient use a scheduling or monitoring tool to execute the run scripts and stop scripts if needed automatically monitor each component s progress and alert support staff if problems arise Scheduling or monitoring tools typically invoke a job control script that executes a Data Ingestion subsystem run and stop scripts In addition using the distributed processing configuration startup approach varies refer to section Destributed Processing Configuration on page 109 for more information Backup Server Configuration An Oracle client can implement a backup server configuration to collect market data in parallel that is in duplicate with the Primary server to help minimize the risk of losing market data for an entire day if the Primary server fails This form of high availability drives configuration of Data Ingestion subsystem components and when to start and stop them In a high availability configuration the backup server transforms and loads market data when the Primary server fails or when market data that the system is collecting on the Primary server is interrupted and causes missing data gaps Also a backup server configuration requires that shared disk be available for checkpoint recovery The daily processing cycle and desired server configuration influences how and when the system starts and stops Data Ingestion subsystem components under normal conditions and if error recovery is necessary Recovery The Data Ingestion compo
341. m creates a new version of the trusted pair record with a status of Risk Esc Rec Cancel RRC However if the party s risk has not increased by enough points to move it to a higher risk zone then no risk review action is initiated on the trusted pair In any case the party s risk will be updated on the applicable Trusted Pair member record The threshold value by which an increase in effective risk will trigger a review of the trusted pair is configurable The default risk zones are configured as RISK_ZONE_1_LOWER 1 RISK_ZONE_1_UPPER 3 RISK_ZONE_2_LOWER 4 RISK_ZONE_2_UPPER 5 SRISK_ZONE_3_LOWER 6 RISK_ZONE_3_UPPER 7 SRISK_ZONE_4 LOWER 8 RISK_ZONE_4_UPPER 10 The ranges of risk values within each zone are configurable but the number of risk zones shall remain at 4 If an implementation chooses not to use all Risk Zones then they can disable them by setting the risk ranges out of bounds For example Risk Zone 1 and Risk Zone 2 may have a lower and upper value of 0 e Customizing Review Reason Text Where the party s effective risk has increased by enough points to move it to a higher risk zone the system also records the reason for marking the record for review This is done using the TP REVIEW REASON TX PARTY1 and TP REVIEW REASON TX PARTY2 parameters Sample strings currently used for review reason text are as follows TP REVIEW REASON TX PARTY1 Recommend Cancel risk of lt Party1 gt increas
342. mal operation Miscellaneous DefaultSourceSystem value Indicates the default value to use for source system when manufacturing reference data records MTS 88 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Ingestion Directory Structure Chapter 4 Data Ingestion Table 32 Datalngest xml File Configuration Parameters Continued Property Name Description Example BufferSize value Specifies the buffer size in kilobytes for I O byte buffers that the MDS and FDT processes create to read input files Use care when changing this parameter due to impact on performance and memory requirements 1024 DirectBufferSize value Specifies the buffer size in kilobytes for Java NIO direct byte buffers that the MDS MDT and FDT processes create to read input files Use care when changing this parameter due to impact on performance and memory requirements 1024 DefaultCurrency value Indicates the value to use as the issuing currency when manufacturing security records from order or trade execution records USD UseDirectBuffers value Specifies whether to make use of Java NIO s direct buffer mechanism Log Specifies properties used to configure the common logging module TRUE Log UseDefaultLog Specifies whether the system uses the default log file for a component The default log file has the name of the compon
343. mantas db pwd in the files SALTIO HOME WEB INF classes hibernate properties and SALTIO HOME WEB INF classes apps NetVis altio app xml For all other Altio apps Change the password for the property db password in the file SALTIO HOME WEB INF classes apps Altio app directory altioapp xml Reports User KDD REPORT OBIEE Reports Open the O0racleBI HOME server Repository and expand the Physical Layer Open the Connection Pool and change the Password parameter to set a new value of the KXDD REPORT schema password Note OBIEE is an optional application Note When you change configuration files under the Alert amp Case Management subsystem you should always make your changes in the original Oracle Financial Services installation directories If you have deployed Behavior Detection as a WAR you must recreate the alert amp case management and admin tools WARs using create am war sh and create at war sh Then re deploy the WAR using your Web Application server console If you are running Behavior Detection directly out of the installation directories you must restart the Behavior Detection web applications using your Web Application server console so that the system recognizes the new passwords Refer to the section on application development in the Oracle Financial Services Bebavior Detection Platform Configuration Guide for more information 48 Oracle Financial Services Behavior Detection Platf
344. mantas sh Changes job status codes from the ERR status to the RES status so that the dispatcher can pick up the jobs with the RES status recover mantas sh Changes job status codes for jobs that were running at the time of a system crash to the ERR status After running this script the restart mantas sh script must be run to change the ERR status code to RES in order for the dis patcher to be able to pick up these jobs In the Oracle Financial Services Job Protocol the processes use a variety of metadata that the database provides Some of this metadata specifies the jobs and their parameters that are associated with the regular operations of an installation Some of this metadata captures the status of job execution and is useful for monitoring the progress of an operational cycle Refer to Chapter 2 of the Oracle Financial Services Behavior Detection Platform FSDM Reference Guide Volume 2 Release 6 1 1 for more information The following sections describe how the processes and metadata interact in the Oracle Financial Services Job Protocol Understanding the Oracle Financial Services Job Protocol Oracle Financial Services Behavior Detection Job templates are maintained through the Scenario Manager These templates associate an algorithm to run with parameters that the algorithm requires Job Templates are grouped together to run in parallel through Job Template Groups in the KDD_JOB_TEMPLATE table Template groups ena
345. mappings between User Roles and sample User Groups are predefined but can also be modified by clients to either adjust the role to sample user group mapping or to map roles to newly defined user groups For mote information about creating a new user group and mapping it to an existing role refer to the Oracle Financial Services Analytical Applications Infrastructure User Manual Release 7 3 Note All User Groups should be mapped to Alert Management Infodom Actions to Role mappings are done through Database tables Sample action to role mappings are included in the application Refer to the Oracle Financial Services Behavior Detection Platform Configuration Guide for more information on changing the mapping of roles to actions Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 27 Setting up a User Chapter 3 Security Configuration Actions are primarily associated with a User Role not an individual user However the ability to Reassign To All when taking a Reassign action is associated at the individual user level Reassign To All means that a user is allowed to assign to users and organizations that may not be within their normal viewing privileges Table 8 describes the predefined User Roles and corresponding User Groups Table 8 Alert Management Roles and User Groups Role Group Name User group Code AM Analyst AM Analyst User Group AMANALYST1GRP AM Analyst II AM Analyst II Us
346. matica workflows the names on transactions are matched against names supplied on watch lists In Scan Watch List Web Service the names provided by a service caller are matched against names supplied on watch lists This Fuzzy Name Matcher utility can be provided directly by Oracle Financial Services software refer to section Fuzzy Name Matcher Utility on page 120 for more information It can also be provided through integration with the Informatica Identity Resolution API Refer to Informatica Identity Resolution documentation for information on installing and the name matching capabilities it provides 126 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Refresh Temporary Table Commands Chapter 4 Data Ingestion Integration between Behavior Detection and Informatica Identity Resolution The names being matched by both the Informatica workflows and the Scan Watch List Web Service ate stored in the WATCH_LIST table Informatica Identity Resolution keys are generated from these names and stored in the SSA_PERSONAL_NAME and SSA_BUSINESS_NAME tables These keys are generated when the Scan Watch List Web Service is started and while it is running through a polling agent and or an Informatica workflow w_ph_load_staging_fuzzy_matches is executed The processes automatically generate these keys you are not required to explicitly run these processes Keys are only generated for names that are not st
347. mation including matches derived from the alerts and actions specified alerts derived from the specified actions and actions derived from the specified alerts However you can run the utility during editing or viewing of other alerts and related information You can also run the utility during alert post processing subject to time constraints Using the Alert Purge Utility The Alert Purge Utility is not patt of an automated batch process that an application such as Maestro or Unicenter AutoSys controls You run this manual process only when necessary Figure 53 The following sections describe configuring and executing the utility as well as the utility s process flow e Configuring the Alert Purge Utility e Executing the Alert Purge Utility e Processing for Purging 216 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Alert Purge Utility Chapter 7 Batch Processing Utilities Configuring the Alert The INSTALL DIR database db tools mantas cfg install cfg file Purge Utility contains common configuration information that the Alert Purge Utility and other utilities require for processing Figure 54 The following sample section from the install cfg file provides configuration information specific to this utility THHHHHHHHHHHHEHHEE ALERT PURGE CONFIGURATION 1HHHHHHEEHBHHEEEHBHBEEEHNI Set the Alert Purge input variables here set the job scenario value you DO NOT USE to null
348. ministration Guide Contents Using the Fuzzy Name Matcher Utility sese ttn ttttntettntentntetnt nte ttetntetn tenen tenentene 123 Configuring the Fuzzy Name Matcher Uifity eee eese estate tete tette tenete tette tenentes 123 Executing the Fuzzy Name Matcher UNH tte toties t PA P Nee ea rie de 126 Using Informatica Identity Resolution API for Name Matching 126 Integration between Behavior Detection and Informatica Identity Resolution eee 127 Refresh Temporary Table Commands 5 ettet edet desi eite cs RE ette tres tei eiie piede tede 127 Use or Control Data itunes dece nee ERE eben be i eite plait eniti eitis 128 Prerequisites for Using Control Data rettet tive rosea rabies iara g aio ran pedo 128 Control Data Inpestotia une oerte tee doe rte Oe e RE A ERREGEA 129 Loading Control Data Thresholds 2 neto a tdeo tate Sun lta Ru wergienss 129 Running Behavior Detection on Control Data w cccccccessseeseeeeesessesesessessesssseesessssssssssssesssssensssesssseessensensenesnss 130 Taboran IN OLS ee cs coL MANDO M CN AR Pene uM MAN a ONU ED NM 130 Resetting the Environments E 131 CHAPTER 5 Informatica Workflows sss 133 About Tnformatica WOEIEIOWS cette te ie dee o iie ase deo otio art cratere 133 Informatica Workflow Types tette ertet iet eroe SEa I EE a etia obses DEEE 134 ZU DEUGCU DL MEM 134 Workflow Tables z
349. mporary Tables 195 Report Tracking Utility 195 Truncate Manager 195 Behavior Detection 3 4 8 Broker Compliance 157 business domain about 36 creating 37 KDD_BUS_DMN table 36 C calendar information 237 Calendar Manager Utility 237 Calendar Manager Utility 195 236 237 calendar information 237 configuring 237 directory structure 236 executing 238 KDD CAL table 238 KDD CAL HOLIDAY table 236 KDD CAL WKLY OFF table 236 logs 236 using 237 Case Management 3 case type subtype 38 about 38 creation 38 usets 31 categories cfg file 208 cloner process 12 shell script 12 component view subsystems 3 tiets 3 configuration files 208 262 263 analysis xml 263 categories cfg 208 DataIngest properties 87 DataIngest xml 62 67 88 DataIngestCustom xml 86 install cfg 198 262 configure 263 Alert Purge Utility 217 analysis XML file 263 Calendar Manager Utility 237 console output utilities 210 constraint element field 264 operator 264 value 264 control data 129 preparation 129 328 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Index correlation alert 172 creation alert 171 case type subtype 38 running alert 174 D Data Analysis Report Error Report 273 Field Distribution Summary Table 273 Null Summary Count Table 273 Query Results 273 Referential Integrity Table Summary 273 SQL Report 273 Table Count Summary
350. ms BusinessSearchLeve 1 SSAParams PersonalMatchLevel Used in defining the type of Search Strategy to use when searching for business names The four possible values allow adjustment to the thoroughness of the search The wider the search the more candidates are typically returned which may increase the reliability of the search however it uses more resources and take longer Used in defining the level of Matching to be performed when searching for personal names The three possible values allow adjustment to the tightness of the match Narrow Typical Exhaustive or Extreme Conservative Typical or Loose 102 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Ingestion Directory Structure Chapter 4 Data Ingestion Table 32 Datalngest xml File Configuration Parameters Continued Property Name Description Example vel SSAParams BusinessMatchLe Used in defining the level of Matching to be Conservative Typical performed when searching for business or Loose names The three possible values allow adjustment to the tightness of the match SSAParams NumberOfQueryOb Specified the number of Informatica Identity 10 jects Resolution sessions open to service requests Each session requires it s own memory area and is used to service single name matching request SSAParams NumberOfInitThr Specifies the number of threads that
351. mum size for an order to 5000 MaxMediumSize be considered medium FDT OrderSizeMarketability Specifies the minimum percent of a small 1 0 SmallMinPercentAtBest order s remaining unit quantity that must be available at the best price for execution to be considered reasonable The minimum percentage value must be represented in its decimal equivalent for example 1 0 100 FDT OrderSizeMarketability Specifies the minimum percent of a medium 1 0 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 95 Data Ingestion Directory Structure Chapter 4 Data Ingestion Table 32 Datalngest xml File Configuration Parameters Continued Property Name Description Example FDT OrderSizeMarketability LargeMinPercentAtBest Specifies the minimum percent of a large order s remaining unit quantity that must be available at the best price for execution to be considered reasonable The minimum percentage value must be represented in its decimal equivalent for example 1 0 100 1 0 FDT TradePurposeFilter value Specifies a comma separated list of trade purpose codes Processing does not consider trades with one of these purpose codes in firm reference price derivations IFADM OFEA CONB CLNT BTBX FDT RunBatchesSeparately value Specifies whether the FDT treats batches as distinct from one another TRUE Three defined batches originate from dif ferent
352. n During the Loading Stage eeiiobe eere tineis riio etes tivi taS AEn E tied tote ebbe tbe E opea ioa ed 112 Data Ingestion Archiving sie entere eo i eben dite cde D Ter HUN T A testa 112 Archiving Database Infotmationiziuio e teen ten mene ROO HRmROSOHeeaHep dones 113 Miscellaneous WU tilt the gis mio ananas pe ba miesciceticciceie tu aa eV EARN 114 Portfolio Manager Utility tret eret rere erento ea e eee e seo ke Posee eoe s ie koe verae eed 116 Change Log Processes cts si ensis escribes he e XY RR XR CERE FPES FART AS HERREN REESE HERI AE fe E 116 Chance oe Summary Process ase te t e t Ea bete e ed p tutes ter ess 118 Network User Account MAD rois oerte er e Se Le ea eq ene eee e Me eno ve ERE E 118 Trade Blotter est tists tat eee aee clas Acad dN theca hse he eee ad ada EEE oun Gottes 118 Trusted Paito fens sete vee de eese do ee een a CLADE Led sedet kd heat 119 Copying Thomson Reference Filesi icssicsssssscssisssssisssdessissadossiessdasssecsesedessstasgseiusedeseseiesesssesedosvsasesessdbessevsdetvsnsebenednsdbesess 119 Copying Other Reference Files retener trt be kl ee RE a EEE NER ER VERS HUE RES sete rua Sede 120 To Copy Other Reference Eales decas temario e Ur rr Rr gRO Fr Me d HS TRUE ERE ives 120 Fuzzy Name Matcher Utility ttd t e Ute ee t E c ice eee oe tava tone t QU oes 120 ID A PR tes 120 viii Oracle Financial Services Behavior Detection Platform 6 1 3 Ad
353. n Case Management Alert Management Web solution am JSPs used in Alert Management Behavior Detection behavior_detection Subsystem Data Ingestion Database Tools ingestion_manager database db_tools Java components Informatica components scripts and stored procedures For DB tools directory Detection Algorithms algorithms C behavior detection algorithms Scenario Manager toolkit Job and scenario editors Financial Services Data Model database Database utilities and database creation scripts Web Services services Web services for watch list scanning and for the alert management supervisor used when posting alerts to Behavior Detection Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide About the Oracle Financial Services Architecture Chapter 1 The Oracle Financial Services Behavior Detection Platform Deployment View Oracle Financial Services architecture from the perspective of its deployment is illustrated in Figure 2 This deployment view illustrates deployment of the major components of each subsystem across servers Additionally the deployment view shows the primary communications links and protocols between the processing nodes Business Data Files and Queues Data Ingestion Server Detection Analyst Oracle Mantas Security Management Workstations Service SMS ry ww Corporate Developer s Intranet Workstation Serve
354. n Guide Scenario Migration Utility Chapter 8 Administrative Utilities Configuring Scenario Extraction Table 108 describes scenario extraction patameters Table 108 Scenario Extraction Parameters Parameter Description extract prod uct range only Flag that indicates the components of the scenario that are shipped with the product value is Y or N the default extract data base username extract data base password User to use to connect to the database when extracting scenarios DB UTIL USER Password for the above user extract conn driver Database connection driver that the utility is to use ora cle jdbc driver OracleDriver extract conn url extract system id Database connection string that the Scenario Migration Utility is to use System from which the scenario was extracted extract schema mantas MANTAS schema owner in the database into which extrac tion of the scenarios occurs MANTAS extract schema busi ness Business schema owner in the database into which extrac tion of the scenarios occurs BUSINESS extract schema market Market schema owner in the database into which extrac tion of the scenarios occurs MARKET extract user miner DATA MINER schema owner in the database into which extraction of the scenarios occurs KDD_MNR extract miner pass word Password for the above user extract dirname Full pa
355. n all these dates run the complete Behavior Detection batch and ingest the data for the respective date Except for Behavior Detection and Post Processing tasks perform all other activities for the Control Data ingestion dates Activities required during any Oracle Financial Services business day are START BATCH gt DRM gt DATA INGESTION gt BEHAVIOR DETECTION gt POST PROCESSING gt END BATCH Prior to running Behavior Detection on the control data you must complete the following procedures 1 Copy all control data from the golden data directory in the database subsystem database golden data directory to the Ingestion Manager inbox directory refer to section inbox Subdirectory on page 106 for more information 2 Run ingestion for all the control data ingestion dates Refer to section Process Flom on page 52 for more information about the ingestion process Note You need to adjust the partitions of the database tables as per the new dates if you intend to process Control Data after the database upgrade to FSDM Loading Control Data Thresholds To generate breaks on the control data specific threshold sets and jobs are created These threshold sets must be installed to the Behavior Detection system for use of control data and generation of test alerts Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 129 Use of Control Data Chapter 4 Data Ingestion 1 Navigate to the direc
356. n page 55 Note A data loader that you configure to run without polling that is batch mode stops automatically when no data remains for processing Running a stopDP sh script does not terminate batch processing runDailySummary sh Launches a process to collect daily summary information runRebuildIndexes sh data type Launches a process to rebuild the indexes of the given component Processing requires this script only during use of a live market feed A valid component value is one of InsideQuote ReportedMarketSale or MarketCenterQuote process firm summary sh Calls a database procedure to build summary statistics about the Oracle client firm data process market summary sh Calls a database procedure to build summary statistics about the Market data market analyze sh Calls a database procedure to create internal database statistics for Market tables firm analyze sh Calls a database procedure to create internal database statistics for Oracle client firm tables Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 83 Data Ingestion Directory Structure Chapter 4 Data Ingestion Table 28 Run or Stop Scripts by Component Continued Script Names Description or Special Instructions runIMC sh Launches the Ingestion Manager Cleaner IMC utility No stop script is available for IMC the utility terminates after it fini
357. nce Events The following sections describe these issues 110 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Rejection During Ingestion Chapter 4 Data Ingestion Lost Events Out of Sequence Events If the system receives an order event other than a New or Cancel or Replace in a set of files before receiving the corresponding New or Cancel or Replace it writes the order event to a lost file The system examines events in the lost file during processing of subsequent sets of files to determine whether the system received the corresponding New or Cancel or Replace event If so processing of this event is normal If an event resides in the lost file when execution of open order processing occurs that is execution of runDP sh OPEN_ORDER processing rejects the event The same applies to trade execution events In addition if a New trade execution event references an otder but the system did not receive the order the New event also resides in the lost file subject to the same rules If rejection of a New or Cancel or Replace order or trade execution occurs during the preprocessor stage all subsequent events ate considered lost events Submission of missing New or Cancel or Replace event can occur in a subsequent set of files and processing of the lost events continue normally An out of sequence event is an order or trade execution event other than New or Cancel or Replace that the s
358. nd Reference data that alert detection processing requires The Ingestion Manager typically receives Market data from a real time Market data feed or file adapter interface and both Business and Reference data through the file adapter interface The Data Ingestion subsystem transforms Market Business and Reference data to create derived attributes that the detection algorithms require The system extracts and transforms data and subsequently loads the data into the database After loading the base tables the Oracle client s job scheduling system invokes Informatica workflows to derive and aggregate data The Informatica component also uses the Fuzzy Name Matcher Utility to compare names found in source data with names in the Watch List The Oracle client implements Ingestion Manager by setting up a batch process that conforms to the general flow that this chapter describes Typically the system uses a job scheduling tool such as Maestro or Unicenter AutoSys to control batch processing of Ingestion Manager Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 51 Process Flow Chapter 4 Data Ingestion Process Flow The Data Ingestion subsystem components receive and process data in a series of workflow steps that include extract or preprocess transform load and post load transformations Figure 11 shows how the Data Ingestion subsystem components participate in each workflow step to extract transform
359. nd in the corresponding XML data map FDT DeriveCorrectionFields Specifies which fields of an order are FDT DeriveCorrectionFields Specifies which fields of a trade are UnitQuantity Trade monitored for corrections PriceIssuing FDT DeriveCorrectionFields Specifies which fields of an execution are UnitQuantity Execution monitored for corrections PriceIssuing FDT DeriveCorrectionFields Specifies which fields of a derived trade are YieldPercentage DerivedTrade monitored for corrections YieldMethodCode LimitPriceIssuing allow before exiting with an error For example a value of 10 allows 10 percent of records to be invalid before exiting with an error A value of 0 allows no invalid records A value of 100 allows all invalid records Order monitored for corrections UnitQuantity XDP Specifies properties used to configure the Preprocessor XDP component XDP Default ArchiveFlag Specifies whether to archive data files The TRUE System copies input files to the backup directory TRUE or deletes input files FALSE XDP Default ErrorLimit Specifies the percentage of invalid records to 100 XDP Default TargetDir Specifies the directory in which to place the resulting output file If this is blank the default output files reside in the corresponding load directory a subdirectory of market load or firm load depending on the schema of the data being processed Oracle Financial Service
360. necessary when a job generates errors and stops during mantas processing If a job stops processing due to errors correct the problems that caused the errors in the job run and restart the job To start a new job execute the restart_mantas sh script with the following parameters restart_mantas sh lt template id gt sd DD MON YYYY ed DD MON YYYY nd It may be necessary to stop one or more job processes when dispatcher errors job errors or some other event make it impossible or impractical to continue processing In addition to stopping the processes administrative intervention may have to resolve the cause of the errors To stop a job you must stop its associated mantas process To obtain the process IDs of active jobs and mantas processes 1 Type ps efw grep mantas and press Enter at the system prompt The mantas processes that are running appear on the computer screen as shown in the following example 00000306 7800 1843 0 Jul 16 ttyiQ iAQM 0 00 kdd datal kdd server bin mantas j 123 The mantas process ID number appears in the first display line in the second column from the left 7800 The job ID number appears in the second display line in the last column j 123 2 Find the job and mantas process ID that you want to stop 3 Type kill mantas process ID at the system prompt and press Enter This command stops the mantas process ID which also stops its associated job Oracle Financial Services Beh
361. nents are designed to be able to restart after a failure Examples of failures include database file system network machine or component After a component fails returns a non zero exit status the general recovery procedure involves checking the component s log file for the cause of the error fix that cause restart the database add more disk space to the file system etc and restart the component using the same command used to start it initially Do not run components that depend on the component that failed until successful completion of the failed component The exception to this procedure is live market Data Ingestion using the Queue Adapter that is when a market feed such as Reuters is in use If a TIBS component ot the machine on which it is running fails recovery of lost data while the component or machine is down is impossible To address this situation Oracle Financial Services recommends that TIBS components be run on two separate machines connected to two separate TIBCO infrastructures If the primary TIBS fails ingestion can proceed with the market data that the backup TIBS produces The client s job scheduling software can be configured to ingest market data through the primary or backup server as needed 108 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Rejection During Ingestion Chapter 4 Data Ingestion Distributed Processing Configuration An Oracle client c
362. nformation class name line number etc cannot be obtained when logging asynchronously Therefore if this information is desired i e specified below the above synchronous properties must be set accordingly for the levels for which this detailed information is desired Also note that this Gb H H o Gt type of detailed information can only be obtained for Java code log format d t p m n Specify the full path and filename of the message library log message library users mantast Solaris10 mantas58 b09 Ti5010S10 Iron 13080 WAS databas e db tools mantas cfg etc mantas database message lib en dat Specify the full path to the categories cfg file log categories file path users mantast Solaris10 mantas58 b09 Ti5010810 Iron 13080 WAS da tabase db tools mantas cfg Specify where a message should get logged for a category for which there is no location property listed above This is also the logging location of the default mantas category unless otherwise specified above Note that if this property is not specified logging will go to the console log default location users mantast Solaris10 mantas58 b09 Ti5010S10 Iron 13080 WAS databa se db tools logs Utilities log Specify the location directory path of the mantaslog if the mantaslog was chosen as the log output location anywhere above Logging will go to the console if mantaslog was selected and this property is not given a value log mantasl
363. ng e Setting Environment Variables e Starting the Dispatcher e Stopping the Dispatcher e Monitoring the Dispatcher Setting Environment Variables Environment variables are set up during the installation process These generally do not require modification thereafter All behavior detection scripts and processes use the system env file to establish their environment About the system env Table 5 describes environment variables in the system env file File Table 5 Environment Variables in system env File Variable Description KDD_HOME Install path of the Oracle Financial Services software KDD_PRODUCT_HOME Install path of the solution set This is a directory under KDD_HOME Table 6 describes database environment variables in the system env file Table 6 Database Environment Variables in system env File Variable Environment Description ORACLE HOME Oracle Identifies the base directory for the Oracle bina ries You must include SORACLE HOMEand ORACLE HOME binin the PATH environment variable value SORACLE HOME lib in the LD LIBRARY PATH environment variable value ORACLE SID Oracle Identifies the default Oracle database ID name to which the application connects TNS ADMIN Oracle Identifies the directory for the Oracle network con nectivity typically specifying the connection infor mation SID Host Port for accessing Oracle databases through SQL NET
364. ng based on the input criteria The utility deletes only alerts and associated actions that exclusively contain matches to be purged Note The system purges matches that do not relate to alerts regardless of the value of limit matches batch size Optional Sets the batch size of purge actions to minimize log space use Specifying a non positive value or specifying no value uses the default of 5 000 rows job Identifies the Behavior Detection Job ID to purge value in the JOB ID column of the KDD JOB table Selecting this variable causes the system to ignore the scenario start date end date andalert status variables Note If you assign a value to the job parameter do not assign a value to the scenario parameter Likewise if you assign a value to scenario assign a value of NULL to job If both the Job ID and the Scenario ID are assigned values the Alert Purge Utility continues to run using the Job ID ignoring the Scenario ID Scenario Identifies the Behavior Detection scenario ID to purge value in the SCNRO ID column of the KDD SCNRO table Note If you assign a value to scenario assign a value of NULL to job Likewise if you assign a value to job assign a value of NULL to scenario If both the Job ID and the Scenario ID are assigned values the Alert Purge Utility continues to run using the Job ID ignoring the Scenario ID 218 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide
365. ng execution of the runFDT sh script Ingestion Manager processes trade related data orders and executions and trades through the Firm Data Transformer or FDT Figure 17 The FDT does the following Enriches data Produces summary records for orders and trades e Calculates derived values to support detection needs e Dertives state chains that is order life cycle states marketability states and displayability states e Provides data for loading into FSDM schema The system executes the FDT with the runFDT sh script the following provides a sample command INSTALL DIR ingestion manager scripts runFDT sh Figure 17 Firm Data Transformer FDT Processing Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 65 Process Flow Chapter 4 Data Ingestion The FDT e Processes all files that reside in the data firm transform directory for the current date and batch e Terminates automatically after processing files that it found at startup e Ignores files that the system adds after processing begins the system may process these files by starting FDT again after exiting from the previous invocation When Ingestion Manager executes runFDT sh it places output files in the directories in Table 22 Table 22 runFDT sh Output Directories Directory Description data firm transform Rollover data that processing saves for the next run of the FDT Includes open
366. nning this utility for the first time after installation execute the command as specified below Note that all passwords need to be entered and it 1s not possible to skip a password INSTALL DIR changePassword sh all For changing password for admin tools subsystem execute the command FIC HOME AM changePassword sh This prompts for the passwords of the following users e Web Application User e Data Miner User When changing a password for the admin tools subsystem if the Web application is deployed from a WAR file the WAR file needs to be regenerated by running FIC HOME AM create at war sh Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 297 Password Manager Utility Chapter 8 Administrative Utilities 298 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide APPENDIX A Logging This appendix describes the mechanism that Oracle Financial Services Behavior Detection Platform uses when logging system messages e About System Log Messages e Message Template Repository e Logging Levels e Logging Message Libraries e Logging Configuration File About System Log Messages The Common Logging component provides a centralized mechanism for logging Behavior Detection messages in which the system places all log messages in a single message library file In the event that a log file becomes very large one gigabyte or more the system creates a new log fil
367. nsists of a Java file that resides in the directory lt INSTALL_DIR gt database db_tools lib and UNIX script files that reside in INSTALL DIR database db tools bin start mantas batch sh starts the batch process end mantas batch sh ends the batch process get mantas batch sh obtains the name of the currently running batch The utility also uses common parameters in the configuration file INSTALL DIR database db tools mantas cfg install cfg Refer to install File on page 198 for more information The following sections describe the Batch Control Utility e Batches in Behavior Detection e Directory Structure e Logs e Using the Batch Control Utility Note To calculate the age in business days versus calendar days verify that the age alerts useBusinessDays setting in the INSTALL DIR database db tools mantas cfg install cfg file has a value of Y yes Batches in Behavior Detection Except for the Alert Management subsystem batches govern all other activity in the Behavior Detection system A batch provides a method of identifying a set of processing This includes all activities associated with Data Ingestion and Behavior Detection Deployment of a system can be with a single batch or with multiple batches You can use multiple batches to permit intra day processing to generate results several times per day or to separate processing based on servicing multiple time zones Behavior Detection provides t
368. ntain all batches in the system When a batch starts as part of an automated process it uses the batch names and other start up information in this table Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 225 Batch Control Utility Chapter 7 Batch Processing Utilities Table 83 provides the contents of the KDD_PRCSNG_BATCH table Refer to the Oracle Financial Services Behavior Detection Platform FSDM Reference Guide Volume 2 for more information about this table Table 83 KDD PRCSNG BATCH Table Contents Column Name Description PRCSNG BATCH NM Name of the batch for example DLY PRCSNG BATCH DSPLY NM Readable name for the batch for example Daily PRCSNG ORDER Relative order of a batch run within processing EOD BATCH NM Name of the batch that is this batch s end of day This name is the same as the name for PRCSNG BATCH NMif the row represents an end of day batch Each row in the KDD_PRCSNG_BATCH table represents a batch Each batch identifies the batch that is the corresponding end of day batch The following three examples illustrate this concept e Single Batch e Single Site Intra day Processing e Multiple Countries Single Batch In this example the KDD_PRCSNG_BATCH table contains a single batch per day This is typical of deployment of a single geography for which a solution set does not require detection more than once daily The KDD_PRCSNG_BATCH
369. ntains additional subdirectories that contain preprocessed and transformed Market data ready for loading into the database Each loader component monitors its assigned subdirectory that is data queue looking for data to load into the database A subdirectory exists for each kind of Market data that a loader moves into the database After loading data files into the database each loader moves the processed files to the backup subdirectory Table 35 identifies the files that each data loader reads and loads into the database Table 35 Files that Market Data Loaders Read and Process Component Input Data Files MDT InsideQuote MDT MDT MarketCenterQuote MDT MDT MarketState MDT MDT ReportedMarketSale MDT Preprocessor data type XDP The backup subdirectory stores files that Data Ingestion subsystem components processed and require no further processing That is they are considered to be in a final form after successful processing e Transformers back up files that they receive and create e Loaders back up files that they finished loading Each file in the backup directoty appears in a subdirectory with the date as its name The name is in the format YYYYMMDD where YYYY is the four digit year MM is the two digit month and DD is the two digit day The IMC component runIMC sh cleans up the backup subdirectory The IMC s configuration file defines the number of days that backup files age before r
370. ntas58 b09 Ti5010810 Iron 13080 WAS databas e db tools data analysis html Valid values for dat output format are HTML and TEXT dat output format HTML Delimiter only applies to TEXT output format dat output delimiter Continued on next page 204 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide About Administrative Utilities Chapter 7 Batch Processing Utilities Continued from previous page HHHHHHHHH Execute Query Tool CONFIGURATION H HHHHHHHHEHHFHEHHHHE Username and password for connecting to the database eqt database username ingest user eqt database password ingest password HHHHHHHHHHHHH Database Builder Utility Configuration H H H H HHHHHH File containing tokens and their value db tools tokenfile users mantast Solaris10 mantas58 b09 Ti5010S10 Iron 13080 WAS database db tools mantas cfg db variables cfg Oracle DuplicateRow 1 Oracle ObjectExists 955 2260 2275 1430 1442 1451 957 1408 2261 Oracle ObjectDoesNotExist 942 1418 1434 2441 904 4043 1927 2443 THHHHHHHHHHHHE Correlation Migration Utility Configuration 4HHHHHHHHBHEBHBHE corrRuleMig CorrRuleFileNm corrRuleMig loadHistory Y aps service url http localhost 8070 mantas services AlertProcessingService HHHHHHHHHHHHH Config Migration Utility Configuration H H HHHHHHHHHHH config filenm prefix Config HHEHHHHHHHHHHHHHHHHH LOG CONFIGURATION 1H jEEUHUEEEHEERRUE D UE D EHE E HH
371. ntifier ALERT ATTR ID Alert Age 113000057 Due Date 113000024 Focus Type 113000010 Last Action 113000038 Owner s Organization 113000056 Previous Match Count All 113000054 Previous Match Count Same Scenario 113000053 Scenario 113000013 Score 113000022 Status 113000008 Status Name 113000055 Processing Batch Name 113000068 Jurisdiction 113000067 Previous Match Count Same Scenario Class 113000064 Scenario Class 113000014 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 179 Auto Close Chapter 6 Post Processing Tasks To View All Alert To view a full set of Alert Closing Attributes run the following query Closing Attributes 1 Select A ATTR_ID A ATTR_NM From KDD ATTR A KDD DATASET ATTR B where A ATTR ID B ATTR ID and B DATASET ID 113000002 Note If the alert attribute that corresponds with a particular alert identifier contains a NULL value the Auto Close algorithm does not interpret these values and returns a fatal Behavior Detection error 2 Formulate operations for the auto closing criteria Operations contain only one mathematical operator for example gt or Operation sets include one or more operations chained together by the NEXT OPRTN SET column 3 Determine an order of precedence for the operations that is what to test first second and so forth Each operation s precedence must be unique within the KDD AUTO CLOSE ALERT table Note An erro
372. ntrol Utility e Setting Up Batches e Starting a Batch Process Manually e Processing for Batch Start e Ending a Batch Process e Processing for End Batch e Identifying a Running Batch Process e Processing for Obtaining a Batch Name The INSTALL DIR database db tools mantas cfg install cfg file contains common configuration information that Batch Control and other utilities require for processing Refer to Figure 54 on page 207 The following sample section from the install cfg file provides configuration information specific to this utility including the single parameter that batch control requires JHHHHHHHHHHHHHHE BATCH CONTROL CONFIGURATION H H H HHHHHHHHHHHHHHHH When ending the batch age alerts in calendar or business days age alerts useBusinessDays Y Figure 60 Configuring Batch Control Utility The value of the age alerts useBusinessDays parameter indicates that at completion of an end of day batch process the Behavior Detection application calculates the age of active alerts by number of calendar days N or business days Y The value of this parameter resides in the KDD_CAL table Refer to Table 93 on page 239 for more information The utility connects to the database employing the user that the utils database username property specifies in the install cfg file The Oracle Financial Services FSDM delivers with a default batch called DLY The KDD PRCSNG BATCH table includes this batch and must co
373. nts a juris diction for example N for North or S for South JRSDCN NM Name of the jurisdiction for example North or South JRSDCN DSPLY NM Display name of the jurisdiction for example North or South JRSDCN DESC TX Description of the jurisdiction for example Northern US or Southern US 2 Add records to the table by using a SOL script similar to the sample script in Figure 6 INSERT INTO KDD JRSDCN JRSDCN CD JRSDCN NM JRSDCN DSPLY NM JRSDCN DESC TX VALUES N North North Northern US Figure 6 Sample SQL Script for Loading KDD JRSDCN Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 35 About Configuring Access Control Metadata Chapter 3 Security Configuration Creating Jurisdiction in Database through Excel Upload Note The KDD_JRSDCN table is empty after system initialization and requires populating before the system can operate The Excel upload process inserts the data into the appropriate dimension tables based on the pre configured Excel upload definitions installed as part of the application installation Data already existing should not be loaded again as this would result in failure of upload When uploading additional records only the incremental records should be maintained in the Excel template with the correct unique identifier key 1 All template Excel files for Excel upload are available in ftpshare STAGE Excelupload
374. o Creating Jurisdiction in tbe Database on page 35 for the steps to perform the Excel Upload of Case Subclass1 The Excel template to be used is KDD CASE SUBCLASS1 xls Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 39 About Configuring Access Control Metadata Chapter 3 Security Configuration Creating Case Subclass2 in Investigation Schema Adding Entries through Excel Upload You can create a Case Subclass2 in the database in the following ways Adding Entries directly in the Table using scripts Adding the appropriate record to the KDD CASE SUBCLASS2 database table Adding records to the table by using a SQL script similar to the following sample sctipt insert into KDD SUBCLASS2 CASE SUBCLASS2 CD CASE SUBCLASS2 NM CASE SUBCLASS1 DESC LAST UPDATED DT LAST UPDATED BY COMMENTS values BSA Bank Secrecy Act Bank Secrecy Act null null null Refer to Creating Jurisdiction in tbe Database on page 35 for the steps to perform the Excel Upload of Case Subclass2 The Excel template to be used is KDD CASE SUBCLASS2 xl1s Creating Case Type and Class Map in Investigation Schema Adding Entries directly in the Table using script Adding Entries through Excel Upload You can create a Case Type and Class Map in the database in the following ways e Add the appropriate record to the KDD TYPE CLASS MAP database table e Add records to the table by using a SOL script sim
375. o data loaded into the MARKET schema MARKET ExtractDir Specifies the parent directory for directories data market extract where the MDS component stores intermediate market data files MARKET TransformDir Specifies the directory where the MDT data market transfor component stores intermediate market data m files MARKET LoadDir Identifies the parent directory for directories data market load that store market data files prior to loading with the Java data loader component Control files for native loaders also reside below this directory BUSINESS Specifies properties related to data loaded into the BUSINESS schema BUSINESS ExtractDir Identifies the parent directory for data firm extract intermediate files that preprocessors produce that are applicable to the BUSINESS schema in the database BUSINESS TransformDir Specifies the working directory for the FDT data firm transform component which transforms BUSINESS trade related data BUSINESS LoadDir Indicates the parent directory for directories data firm load that store BUSINESS schema bound data files prior to loading with the Java data loader component Control files for native loaders also reside below this directory MANTAS Specifies properties related to data loaded into the MANTAS schema MANTAS ExtractDir Specifies the parent directory for data mantas extract intermediate files that
376. o running AML scenarios These scripts refresh the required temporary tables for selected AML scenario detection Refer to section Refreshing Temporary Tables on page 255 for more information Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 127 Use of Control Data Chapter 4 Data Ingestion Use of Control Data After installing the Oracle Financial Services software you can use control data that Oracle Financial Services Behavior Detection Platform provides to test end to end processing of data that is running data ingestion executing scenarios and viewing generated alerts in the Alert Management UI This control data provides data that runs with the scenarios that you have purchased Thus you can verify that installation of the software is correct and works as designed To prepare the system for testing follow these steps 1 Complete the prerequisites for using control data refer to section Prerequisites for Using Control Data on page 128 for more information 2 Prepare for ingestion of the control data refer to section Control Data Ingestion on page 129 for more information 3 Install the control data refer to section Loading Control Data Thresholds on page 129 for more information 4 Run Behavior Detection on control data to generate alerts refer to section Running Behavior Detection on Control Data on page 130 for more information Prerequisites for Using Control
377. o stop the service These scripts can be found in the INSTALL_DIR services scripts directory Details of the scripts are as follows startWebServices sh Run this script to start the web service shutdownWebServices sh Run this script to stop the web service 296 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Password Manager Utility Chapter 8 Administrative Utilities Password Manager Utility To change a password in any subsystem other than alert management and admin tools execute the command XINSTALL DIR changePassword sh This prompts for the passwords of all the required application users The passwords that are entered are not output to that is not shown on the screen and the same password needs to be re entered in order to be accepted If it is not necessaty to change a given password press the Enter key to skip to the next password The passwotd that was skipped was not changed The following are the users for which the script prompts for passwords depending on what subsystems have been installed e Data Ingest User e Database Utility User e Algorithm User e Data Miner User If there is a need to change a specific password property stored in an application configuration file the following command can be run INSTALL DIR changePasswords sh property name gt For example INSTALL DIR changePasswords sh email smtp password Note If you are ru
378. oad 52 Post Processing about 171 alert creation 176 alett scoring 176 alert suppression 183 assignment 177 auto close 178 179 Automatic Alert Closure 171 Highlight Generation 184 match scoring 173 Post Watch List AML Banking 150 AML Brokerage 150 Broker Compliance 157 Post Watch List Miscellaneous AML Brokerage 140 Fraud Detection 163 Insurance 169 Post Watch List Miscellaneous workflows AML Brokerage 140 Fraud Detection 163 Insurance 169 Preprocess Workflow 52 preprocessing 57 data type 109 missing data 110 output 61 process flow 60 Pre Watch List Miscellaneous Fraud Detection 159 Pre Watch List Miscellaneous workflows AML Brokerage 137 pre watch list miscellaneous workflows Fraud Detection 159 process flow preprocessing 60 property log format 308 property value console 307 eventviewer 307 mantaslog 307 syslog 307 Push E mail Notification using 255 256 257 utilities 196 332 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Index R record handling lost events 111 out of sequence events 111 record log messages 307 recovering jobs system crash 24 Reference Files Thomson 119 Refreshing Temporary Tables 127 255 logs 255 using 255 utilities 196 Report Framework 313 prerequisites 313 Report Management 313 Report Tracking Utility utilities 196 Restarting 20 roles database 46 Oracle Financial Services installer x
379. ocess DATA DUMP DT Business day on which the batch ran START TS Time that the batch started END TS Time that the batch ended if applicable STATUS CD Status code that indicates whether the batch is cur rently running RUN or has finished FIN 4 The Batch Control Utility logs a message in the INSTALL DIR database db tools logs batch control log file stating that the batch process has begun Querying the KDD PRCSNG BATCH HIST table for confirmation that the batch has started displays information similar to that in Figure 61 In the last entry note the appearance of RUN for STATUS CD and lack of end time in END TS PRCSNG BATCH ID PRCSNG BATCH NM DATA DUMP DT START TS END TS STATUS CD 1 DLY 10 Nov 06 11 Nov 06 11 Nov 06 FIN 2 DLY 11 Nov 06 12 Nov 06 12 Nov 06 FIN 3 DLY 12 Nov 06 13 Nov 06 13 Nov 06 FIN 4 DLY 13 Nov 06 14 Nov 06 14 Nov 06 FIN 5 DLY 14 Nov 06 15 Nov 06 15 Nov 06 FIN 6 DLY 15 Nov 06 16 Nov 06 16 Nov 06 FIN 7 DUY 16 Nov 06 17 Nov 06 17 Nov 06 FIN 8 DLY 17 Nov 06 18 Nov 06 18 Nov 06 FIN 9 DLY 18 Nov 06 19 Nov 06 19 Nov 06 FIN 10 DLY 19 Nov 06 20 Nov 06 20 Nov 06 FIN 11 DLY 20 Nov 06 21 Nov 06 RUN 228 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Batch Control Utility Chapter 7 Batch Processing Utilities Figure 61 Sample KDD PRCSNG BATCH HIST Table Batch Start Status Ending
380. ocessing of some wotkflows requires data from Miscellaneous predecessor workflows 3130 and 3210 refer to Informatica Miscellaneous Optional Workflows AML Banking on page 146 for more information These workflows may proceed simultaneously in all sequences as soon as processing of any predecessors has completed Refer to the note in the section Informatica Miscellaneous Post Watch List Workflows AML Brokerage on page 140 for more information Table 54 describes the Informatica Miscellaneous Post Watch List workflows in Figure 37 Table 54 Post Watch List Workflows AML Banking Sequence Number Folder Workflow Number Workflow Name Predecessor 13 MLM Banking Com 2090 w ph2090 update jurisdiction in cb MLM Banking Common 3130 mon 13 ORION Common 2200 w ph2200 build trxn tables from fotps ORION Common 3051 ORION Common 3210 14 ORION Common 2220 w ph2220 update fot reversals ORION Common 2200 20 ORION Common 3501 w ph3501 Exp and Risk Review TP ORION Common 2200 21 ORION Common 3502 w ph3502 Flag Trusted Trxn ORION Common 3501 150 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide AML Banking Informatica Workflows Chapter 5 Informatica Workflows Informatica Summary Workflows AML Banking Informatica Summary workflows maintain monthly aggregations of customer activity Figure 38 illustrates Informatica Summary workflows
381. og location users mantast Solaris10 mantas58 b09 Ti5010S10 Iron 13080 WAS data base db tools logs mantaslog log Continued on next page 206 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide About Administrative Utilities Chapter 7 Batch Processing Utilities Continued from previous page Specify the hostname of syslog if syslog was chosen as the log output location anywhere above Logging will go to the console if syslog was selected and this property is not given a value log syslog hostname Specify the hostname of the SMTP server if an e mail address was chosen as the log output location anywhere above Logging will go to the console if an e mail address was selected and this property is not given a value log smtp hostname Specify the maxfile size of a logfile before the log messages get rolled to a new file measured in MBs If this property is not specified the default of 10 MB will be used log max size NOTE The values for the following variables need not be changed Specify the ID range for wrapper datasets dataset wrapper range min 113000001 dataset wrapper range max 114000000 product id range min 113000000 product id range max 200000000 Figure 54 Sample install cfg File Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 207 About Administrative Utilities Chapter 7 Batch Processing Utilities
382. olumn gt lt JOIN gt lt Optional May contain one or more filters gt lt FILTER table table name column column operator operator value filter value gt lt MULTIJOIN gt 268 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Analysis Tool Chapter 8 Administrative Utilities The lt FILTER gt element is optional in the join analysis Multiple filters can be applied to a join The AND operator is appended to each filter condition upon creation of the query The following XML code illustrates the use of a filter with a simple join analysis ANALYSIS lt JOINS gt lt MULTIJOIN gt lt JOIN gt LEFT table ACCT column ACCT INTRL ID gt RIGHT table CUST ACCT column ACCT INTRL ID gt lt JOIN gt lt FILTER table ACCT column DATA DUMP_DT operator GTE value 01 NOV 2006 gt lt FILTER table ACCT column DATA DUMP_DT operator LTE value 05 NOV 2006 gt lt MULTIJOIN gt lt JOINS gt lt ANALYSIS gt This code executes the following query select count 1 from ACCT a CUST_ACCT b where a ACCT INTRL ID b ACCT INTRL ID and a DATA DUMP DT 01 NOV 2006 and a DATA DUMP DT 05 NOV 2006 To filter for values that are null or not null set the operator to EMPTY and the value to IS NULL Of IS NOT NULL respectively Join Count by Distinct Column To determine a join count of the number of distinct values for a specified
383. on status code If an application component terminates successfully a script returns a zero return code If the component fails to terminate successfully the script returns a non zero status normally 1 Table 28 defines the run scripts for starting and stopping each component and any special instructions Table 28 Run or Stop Scripts by Component Script Names Description or Special Instructions runTIBS sh stopTIBS sh Launches the TibSpool TIBS through the runTIBS sh script or terminates it through the stopTIBS sh script Processing configures a symbol range for example A C for each instance of TIBS This script uses the range to select the Market data that the TIBS captures You can modify a section within the script that sets the level of detail logged from processing this script The default is a level of 7 which is appropriate for normal processing Use levels 2 and 3 for more detailed diagnostic logging The command runTIBS sh 1 2 runs TibSpool to process market data range 1 which Datalngest xml defines with a log level of 2 Market data ranges are set in DataIngest xml refer to Data Ingest XML Configuration File on page 88 for more information runTIBW sh stopTIBW sh Launches the Tibco Watcher TIBW through the runTIBW sh script or terminates it through the stopTIBW sh script You can modify a section within the script that sets the level of detail logged from processing this script The logging le
384. order 1px solid 000 border collapse collapse Continued on next page Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 203 About Administrative Utilities Chapter 7 Batch Processing Utilities Continued from previous page email style th font style bold border l1px solid 000 border collapse collapse padding 4px background C7DAED email le tr font size 10pt email le td border 1px solid 000 border collapse collapse padding 4px email le footer font family Arial Helvetica sans serif font size 10pt color black email le disclaimer font style italic HHHHHHHHH HIGHLIGHTS GENERATION CONFIGURATION H H HHHHHHHHHHHHHHHHHEH Set the default currency code See mantas cfg etc xml CUR Currencies xml for supported currency codes currency default USD HEHHHHHHH HDC CONFIGURATION JHHHBHUBHBHBHBBHBHBHHEBHIEE Set the maximum number of hdc threads hdc maxthreads 1 hdc batchsize 10000 HHHHHHHHH Data Analysis Tool CONFIGURATION 1HHHEHEEBEEEEEEE EE B ER EE Username and password for connecting to the database dat database username ingest user dat database password ingest password Input file for analysis dat analysis input users mantast Solaris10_mantas58_b09 Ti5010S10 Iron 13080 WAS database db tools mantas cfg analysis aml xml Output file and file format control dat analysis output users mantast Solaris10 ma
385. ored in the SSA tables The inactive names on watch list are not removed however these names are be filtered out and not matched An excessive number of these inactive names can cause performance and storage issues You can remove them by truncating SSA_PERSONAL_NAME and SSA_BUSINESS_NAME tables You can use the following command to truncate the tables truncate table sh SSA PERSONAL NAME and truncate table sh SSA BUSINESS NAME These tables are refreshed from the WATCH LIST table using either the Scan Watch List Web Service ot the Informatica workflows Refer to the Oracle Financial Services Behavior Detection Platform Installation Guide for information on configuring the location of the Informatica Identity Resolution software when installing Oracle Financial Services In addition the value of the WatchListService NameMatcher parameter in DataIngest xml must be changed to SSANameMat cher This instructs the application to use the Informatica Identity Resolution integration when doing name matching in both the Informatica workflows and the Scan Watch List Web Service DataIngest xml also contains three parameters that affect how name matching is done with Informatica Identity Resolution These ate Key Level Search Level and Match Level Collectively they control the names returned when matching a given name Refresh Temporary Table Commands Prior to running post processing you must execute database scripts after ingestion and prior t
386. orm 6 1 3 Administration Guide About Configuring File Type Extensions Chapter 3 Security Configuration About Configuring File Type Extensions The list of file type extensions that are allowed to be attached while performing document attachment action should be configured as comma separated values in CONFIGURATION table of OFSAAT configuration schema in its PARAMVALUE column where PARAMNAME is DOCUMENT_ALLOWED_EXTENSION Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 49 About Configuring File Type Extensions Chapter 3 Security Configuration 50 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide CHAPTER 4 Data Ingestion This chapter discusses the operation of the Oracle Financial Services Data Ingestion processor Ingestion Manager and subsystem components Specifically this chapter focuses on the following topics e About Data Ingestion e Process Flow e Intra Day Ingestion Processing e Informatica Ingestion Parameters e Alternatives to Standard Data Ingestion Practices e Data Ingestion Directory Structure e Startup and Shutdown e Data Rejection During Ingestion e Data Ingestion Archiving e Miscellaneous Utilities e Copying Thomson Reference Files e Fuzzy Name Matcher Utility e Refresh Temporary Table Commands e Use of Control Data About Data Ingestion The Ingestion Manager receives transforms and loads Market Business a
387. ormatica Summary Generation Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 71 Process Flow Chapter 4 Data Ingestion Figure 22 illustrates Informatica Watch List processing and risk assignment Figure 22 Informatica Watch List Processing Refer to section Alternatives to Standard Data Ingestion Practices on page 77 for more information about Watch List processing 72 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Intra Day Ingestion Processing Chapter 4 Data Ingestion Intra Day Ingestion Processing Figure 23 provides a high level flow of the intra day ingestion process of extracting transforming and loading data In ion Pri ing Flow Intr Description TCM processing only using 1 n number of intra day batches data drops Optional Run the firmanalyze sh and marketanalyze sh scripts to improve R intra day detection performance Start Data Ingestion Extract copy DIS formated data from third party system for Reference data and Intra day Batch 1 e g filename _ONE_ dat to the inbox directory Except OpenOrder and CorporateAction Run Beginning of day processes Run Intra day batch ONE Extract copy DIS formatted data from third party system for Intra day Batch 2 to inbox run firmanalyze sh marketanalyze sh Run detection for BEX and some TC s
388. ote a correlation to a case Both actions have associated parameters that are defined and dictated by the rule that generated the correlation these rule sets are discussed below Note that the promote fo case action is also a licensable feature dependent on Integrated Case Management license The same information as above applies in terms of obtaining and configuring a license file Both parameters above can be configured by changing their associated VALUE TX values in the KDD_ PARAM BINDING table In addition to the job parameters there is a certain metadata that needs to be in place in order to successfully run Alert Correlation These include the definitions of the paths used to correlate alerts to business entities and the correlation rules that define the criteria for correlating alerts to alerts and the parameters associated to any subsequent actions performed if this step in the process is chosen to be run Details on this metadata is provided in the following sub sections The business entity paths are currently managed through manual interaction with the KDD BUS NTITY PATH and KDD BUS NTITY PATH CFG tables in the FSDM These tables are populated with a comprehensive set of sample data paths However the following information will assist in modifying these paths or adding to them The structure of the tables is as follows Table 74 Table 74 KDD BUS NTITY PATH Metadata Table Column Name Primary Key Foreign Key Col
389. ough the Behavior Detection engine log file when it runs in trace mode Using the Get Dataset Query With Thresholds Utility Processing extracts the dataset query and uses it as input for tuning and execution plan generation Note This utility does not recursively substitute thresholds in child datasets Therefore if a dataset being extracted has a reference to another dataset manual extraction of that dataset must also occur Table 106 describes the parameters to provide with the get_dataset_query sh script Table 106 Get Dataset Query Variables Parameter Description Dataset ID Unique identifier of the dataset for retrieval Threshold Set Unique identifier of the threshold set for retrieval ID Executing the Get Dataset Query with Thresholds Utility The following section provides instructions to execute the Get Dataset Query with Thresholds Utility To Execute the Get To execute the Get Dataset Query with Thresholds Utility follow the steps Dataset Query with Thresholds 1 After the Alert Creator process completes execute the get_dataset_query sh script as follows INSTALL DIR database db tools bin get dataset query sh lt Data set ID Threshold Set ID The dataset query automatically prints to standard output which you can copy and paste into any other application When the dataset query does not find a dataset output is Error Dataset not found When the dataset query does not fin
390. ows may proceed simultaneously in all sequences as soon as processing of any predecessors has completed Table 69 details the Watch List workflows for Insurance in Figure 52 Table 69 Summary Workflows Insurance Sequence Number Folder Workflow Number Workflow Name Predecessor 16 ORION_Production 0554 w_ph0554_truncate_IPSS 17 ORION Production 0550 w ph0550 aggregate IPSS from IT ORION_Production 0554 18 ORION Production 0551 w ph0551 IPSS from IPB ORION_Production 0550 19 ORION_Production 0552 w_ph0552_build_IPSD_from_IPSS ORION_Production 0551 19 ORION Production 0553 w ph0553 build IPSM from IPSS ORION Production 0551 170 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide CHAPTER 6 Post Processing Tasks This chapter defines the following post processing administrative tasks About Post Processing About Post Processing Match Scoring Alert Creation Update Alert Financial Data Alert Scoring Assignment Auto Close Automatic Alert Suppression Highlight Generation Augment Trade Blotter Score Trade Blotter Historical Data Copy Alert Correlation Account Approval Pre Trade Approval Tasks During post processing of ingested data Behavior Detection prepares the detection results for presentation to users Preparation of the results depends upon the following processes Augmentation Collects information for pattern detection which
391. percentage values Table 42 Fuzzy Name Matcher Utility Configuration Parameters Parameter Description fuzzy name stopword file Identifies the file that stores the stop word list The stop word file is either corporate or personal The lt prefix gt token identifies corporate as B and personal as P Certain words such as Corp Inc Mr Mrs or the do not add value when comparing names fuzzy name match threshold Indicates the score above which two names are considered to match each other The utility uses this parameter only when the match multi property has a value of true The allowable range is from 0 to 100 fuzzy name initial match score Specifies the score given for matching to an initial The allowable range is 0 to 100 the recommended default is 75 fuzzy name fuzzy name initial match pl initial match p2 Specifies the number of token picks that must be made before awarding initial match score The value is an integer gt 0 The default value is 2 Specifies the number of token picks that must be made before awarding initial match score if only initials remain in one name The value is an integer gt 0 The default value is 1 fuzzy name extra token match score Indicates the score given to extra tokens The allowable range is O to 100 the recommended default is 50 fuzzy name extra token min match Specifies the minimum number of matches that occur
392. pically an Oracle client uses the default name of MANTAS Data Ingestion accesses the MANTAS schema when allocating sequence IDs to ingested records MANTAS MARKET DBSchema Schema name for the MARKET database schema Typically an Oracle client uses the default name of MARKET Data Ingestion stores market data related records in the MARKET schema MARKET BUSINESS DBSchema Schema name for the BUSINESS database schema Typically an Oracle client uses the default name of BUSINESS Data Ingestion stores market data related records in the BUSINESS schema BUSINESS TargetDir XDP AccountProfitAndLoss Name of the source files directory for the Data Ingestion informatica installation Java ingestion places some files that Informatica mappings require into this directory software informatica PC 811 SrcFiles Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 87 Data Ingestion Directory Structure Chapter 4 Data Ingestion Table 31 Datalngest properties File Configuration Parameters Continued Property Name Description Example MDS Adapter RvSession Service Service name for the TIBCO live market feed Only Oracle clients who opt to use the queue adapter to process live market data use this parameter 7602 MDS Adapter RvSession Network Network name for the TIBCO live market feed Only Oracle clients
393. ppends its extension to the end of the file Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 103 Data Ingestion Directory Structure Chapter 4 Data Ingestion data market Subdirectory extract Subdirectory transform Subdirectory Table 33 identifies the error file signatures that each component can output to the errors subdirectory Table 33 Error File Signatures Output by Component Component Error File Preprocessor data type XDP err Data Loader data type XDL err FDT Order FDT err TradeExecution FDT err MDS InsideQuote MDS err MarketCenterQuote MDS err ReportedMarketSale MDS err The IMC utility zunIMC sh cleans up the errors subdirectory The IMC s configuration file defines the number of days that error files age before their removal The market subdirectory within the data subdirectory contains the extract transform and load subdirectories that correspond directly to the workflow steps that market data moves through during Data Ingestion The following subsections describe each subdirectory in more detail The extract subdirectory within the market subdirectory contains additional subdirectories that organize preprocessed Market data It organizes data by date that is YYYYMMDD where YYYY is the four digit year MM is the two digit month and DD is the two digit day The MDS extracts and preprocesses market data that con
394. ppression optional Suppresses alerts that share specific scenario and focal entity attributes for a particular time frame Refer to Automatic Alert Suppression on page 183 for more information Highlight Generation Generates highlights for alerts that appear in the alert list in the Alert Management subsystem and stores them in the database Refer to Highlight Generation on page 184 for more information Augment Trade Blotter Provides the ability to differentiate between various types of trades using text based codes It also provides the ability to flag trades that require additional analysis before an analyst can mark trade as Reviewed or Reviewed with Follow up Refer to Augment Trade Blotter on page 184 for more information Score Trade Blotter Determines the maximum score of alerts generated in the same batch cycle associated with a trade also determines the alert trade mappings Refer to Score Trade Blotter on page 185 for more information Historical Data Copy Identifies the records against which the current batch s scenario runs generated alerts and copies them to archive tables Refer to Historical Data Copy on page 185 for more information Alert Correlation Uncovers relationships among alerts by correlating alerts to business entities and subsequently correlating alerts to each other based on these business entities this latter step is optional The relationships are discovered based on configurable rule sets R
395. preprocessors produce that are applicable to the MANTAS schema in the database MANTAS TransformDir Specifies the working directory for data mantas transfor intermediate files that utilities produce that m are applicable to the MANTAS schema in the database MANTAS LoadDir Specifies the parent directory for directories data mantas load that store MANTAS schema bound data files prior to loading with the Java data loader component Control files for native loaders also reside below this directory Directory Specifies properties used to define directory locations 90 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Ingestion Directory Structure Chapter 4 Data Ingestion Table 32 Datalngest xml File Configuration Parameters Continued Property Name Description Example Directory Log Specifies the parent directory for log file directories and log files that Java ingestion components create logs Directory Inbox Specifies the input directory where Java ingestion components find files that the Oracle client submits Processing creates subdirectories in the inbox directory for each day of data to contain a copy of the input data file inbox Directory Error Specifies the parent directory for error directories that contain error data files that Java ingestion components create Each error data file contains records
396. processing of any predecessors has completed Table 63 details the Informatica Watch List workflows in Figure 46 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 161 Fraud Detection Informatica Workflows Chapter 5 Informatica Workflows Table 63 Watch List Workflows Fraud Detection Sequence Workflow Number Folder Number Workflow Name Predecessor 3 ORION Common 3000 w ph3000 Adjust WL WLS ORION Common 2180 3 ORION Common 3005 w ph3005 apply cust KYC risk 3 ORION Common 3030 w ph3030 truncate wls 4 ORION Common 3010 w ph3010 truncate Is 4 ORION Common 3020 w ph3020 truncate nms 4 ORION Common 3040 w ph3040 truncate wls2 4 MLM Brokerage C 3041 w ph3041 create addresses from inst ommon ORION Common 13051 w ph3051 create addresses from fotps 4 ORION Common 3090 w ph3090 create external entities from fotps ORION Common 2190 ORION Common 3091 4 MLM Brokerage C 3120 w ph3120 create external entities from inst MLM Brokerage Common 2192 ommon 4 MLM Banking Com 3130 w ph3130 create client banks from fotps ORION Common 2190 mon 5 ORION Common 3070 w ph3070 load watch list staging table ORION Common 3000 ORION_Common 3030 5 ORION Common 3140 w ph3140 write fotps associations to ls ORION Common 3010 ORION Common 3051 ORIO
397. put on page 210 for more information Processing for Batch After establishing the required Java environment and initiating various Java processing Start activities the Batch Control Utility does the following 1 The utility verifies that the provided batch name contains only the characters A Z a z and 0 9 by querying the KDD_PRCSNG_BATCH table Table 86 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 227 Batch Control Utility Chapter 7 Batch Processing Utilities 2 The utility determines whether a batch is running by querying the KDD_PRCSNG_BATCH_CONTROL table Table 87 Table 87 KDD PRCSNG BATCH CONTROL Table Contents Column Name Description PRCSNG BATCH ID Current batch process ID PRCSNG BATCH NM Name of the current batch process DATA DUMP DT Current business day The Calendar Manager Utility places this information in the table EOD PRCSNG BATCH FL Flag that indicates whether the batch is an end of day process Y or N 3 The utility records information about the batch in the KDD_PRCSNG_BATCH_HIST table This table contains a history of all batches that appear by start date and end date Table 88 describes the KDD PRCSNG BATCH HIST table Table 88 KDD PRCSNG BATCH HIST Table Contents Column Name Description PRCSNG BATCH ID Current batch process ID PRCSNG BATCH NM Name of the current batch pr
398. r Alert amp Case Management Behavior Data Ingestion Figure 2 Oracle Financial Services Architecture Deployment View The complex interactions between the components of the Alert amp Case Management subsystem become apparent in the deployment view The Alert amp Case Management subsystem requires the following e Web browser e Web server e Web Application server e Security Management Service SMS Oracle Financial Services Alert Management and Case Management use inbuilt SMS Security Management Service for handling both authentication and authotization The Alert amp Case Management subsystem also supports the use of an External Authentication Management EAM tool to perform user authentication at the Web servet if a customer requires it Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 5 About the Oracle Financial Services Architecture Chapter 1 The Oracle Financial Services Behavior Detection Platform These components can operate when deployed on a single computer or when distributed across multiple computers Figure 2 In addition to being horizontally scalable the application is vertically scalable in that replication of each of the components can occur across multiple servers Security View The security view of the architecture and use of security features of the network in a Behavior Detection architecture deployment is illustrated in Figure 3 Behavior
399. r It updates Customer Balance position table 2020 w p2020 delete nonstanding instruc This workflow can be used to delete old Instructions tions 2040 w p2040 update cust count of accou This workflow calculates the total number of accounts nts for an institutional customer ORION Common Folder 2005 w p2005 set cd pd dates This workflow sets the CD Date and PD Date values to the CDPDdate file The post session success command updates the Parame ter file with CD Date and PD Date entries 2050 w ph2050 update bot reversals This Workflow handles reverserals for Back Office Transactions 2070 w ph2070 truncate cas The workflow truncates CUST ACCT STAGE 2100 w ph2100 truncate atps This workflow truncates ACCT TRD POSN STAGE 2110 w ph2110 update bot unrelated part This workflow populates the unrelated party code for y code BOT 2120 w ph2120 load ATPS from trade This workflow populates trades into a staging table for use later in calculating open close position and day trades All non canceled trades for a day are inserted 2130 w ph2130 update fot unrelated part This workflow populates the unrelated party code for y code FOT 2140 w ph2140 pass thru process This workflow kicks off the Pass Thru process It gener ates second originator and beneficiary records for Front Office Transaction It also sets the pass thru flag based on the a set of expressions Or
400. r you can customize these tables The utility accommodates daily weekly and monthly partitioning schemes It also processes specially configured Mixed Date partitioned tables The Mixed Date tables include partitions for Current Day Previous Day Last Day of Week for weeks between Current Day and Last Day of Previous Month and Last Business Day of Previous Two Months The Data Retention Manager can e Perform any necessary database maintenance activities such as rebuilding global indexes e Add and drop partitions or both to or from the date partitioned tables Data Retention Manager provides a set of SQL procedures and process tables in the Behavior Detection database A shell script and a configuration file that contain the various inputs set the environment that the utility uses This section covers the following topics e Directory Structure e Logs e Processing Flow e Using the Data Retention Manager e Utility Work Tables Table 94 provides the directory structure for the Data Retention Manager Table 94 Data Retention Manager Directory Structure Directory Contents bin Executable files including the xzun drm utility sh shell script lib Required class files in jar format mantas cfg Configuration files for example install cfg and categories cfg in which you can configure properties and log ging attributes logs File INSTALL DIR database db tools logs DRM Utility log thatthe util
401. r Detection Platform 6 1 3 Administration Guide Setting up a User Chapter 3 Security Configuration e Organization s Department or organization to which an individual user belongs e Role s Permissions or authorizations assigned to a user in the system such as Oracle Financial Services administrator or Auditor e Scenario Group s Group of scenarios in Oracle Financial Services that identifies a set of scenario permissions and to which a user has access rights applicable to Alert Management only e Case Type Subtype s Case type subtypes combinations to which a user has access rights applicable to Case Management only Figure 5 illustrates the Oracle Financial Setvices user authorization model What data types What is the scope can the user of entities that the view user monitors Request Alerts amp Research FSDM Business Data Alerts Response ulewog ssauisng e Uu D a E e uU CE zt UO D What functions can the user What types of perform Cases can the user work on With what business problems does the user work How are the users organized Figure 5 Oracle Financial Services User Authorization Model Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 31 Setting up a User Chapter 3 Security Configuration Table 11 provides the relationships between the data points that Figure 5 illustrates Table 11 Relationships between
402. r Workflow Number Workflow Name Predecessor 1 ORION Common 2005 w p2005 set cd pd dates 1 ORION Common 2050 w ph2050 update bot reversals 1 ORION Common 2070 w ph2070 truncate cas 1 ORION Common 2180 w ph2180 instn identification 1 ORION Common 2450 w ph2450 populate ACCT SRVC TEAM 2 ORION Common 2190 w ph2190 fotps instn processing ORION Common 2180 2 ORION Common 2130 w ph2130 update fot unrelated party code ORION Common 2050 ORION Common 2070 154 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Broker Compliance Informatica Workflows Chapter 5 Informatica Workflows Informatica Summary Workflows Broker Compliance Informatica Summary workflows maintain monthly aggregations of customer activity Figure 41 illustrates Informatica Summary workflows for BC External Predecessors 2010 R lsc isc isc Misc 0171 0370 0400 0410 040 0420 0350 Sequence 17 0390 Sequence 16 Sequence 18 Cost Sequence 19 0630 0640 Sequence 20 Note Processing of wor OWS IS no sequence dependent processing can occur simultaneously across all sequences Sequence 21 q Workflows with predecessors must wait for predecessor results before workflow processing can begin Figure 41 Summary Workflows Broker Compliance The application completes these workflows in five sequences processing of some workflows requires data from exte
403. r datasets need to be extracted or loaded score include N wrapper include N Specify the Use Code for the scenario Possible values are BRK or EXP load scnro use BRK Specify the full path of depfile and name of fixfile used for extraction and loading Note fixfile need not be specified in case of loading sm depfile users oriont mantas6 1 3 database db tools mantas cfg dep cfg sm release 6 1 3 HHHH EXTRACT Specify the database details for extraction extract database username utils database username extract database password utils database password Specify the jdbc driver details for connecting to the source database extract conn driver database driverName extract conn url jdbc oracle oci T209S8 Source System Id extract system id TEST ENVIORNMENT Specify the schema names for Extract extract schema mantas schema mantas owner extract schema business schema business owner extract schema market schema market owner extract user miner load user miner extract miner password utils miner password File Paths for Extract Specify the full path in which to place extracted scenarios extract dirname users oriont mantas6 1 3 database db tools data Specify the full path of the directory where the backups for the extracted scripts would be maintained extract backup dir users oriont mantas6 1 3 database db tools data temp Continued on next page 276 Oracle Finan
404. r occurs if two operations have the same precedence All operations must have a precedence or the system does not test them 4 Assign an operation ID to each operation This ID must be unique within KDD AUTO CLOSE ALERT 5 Assign an operation ID to each operation within each operation set Use IDs close together for operations within the same operation set The system uses this ID to link together operations within the same operation set by placing the next ID for testing in the Next Operation ID NEXT OPRTN SET ID col umn 6 Determine the rows to insert into the KDD AUTO CLOSE ALERT table from the following columns OPRTN SET ID is the operation set ID XPRSN ORDER ID the operation ID the precedence must be unique for each operation across the table This column can contain a NULL value Note When an operation set is reached by linking from another operation set you can leave the XPRSN ORDER ID at NULL For operations sets that are not reached through another operation set the XPRSN ORDER ID is required ALERT ATTR ID Refer to Step 1 OPRTR CDis the mathematical operator for the operation VALUE TXis the right hand side of the operation NEXT OPRTN SET IDis the ID that identifies the next operation in the operation set or NULL if no operations exist Note Inserting an ID into the NEXT OPRTN SET column previously called creates a loop and results in an error 180 Oracle Financial Services
405. r record to the KDD_BUS_DMN database table which Table 14 describes Table 14 KDD_BUS_DMN Table Attributes Column Name Description BUS DMN CD Single character code that represents a business domain for example a b or c BUS DMN DESC TX Description of the business domain for example Insti tutional Broker Dealer or Retail Banking BUS DMN DSPLY NM Display name of the business domain for example INST or RET MANTAS DMN FL Flag that indicates whether Behavior Detection speci fied the business domain Y If an Oracle client speci fied the business domain you should set the flag to N The KDD BUS DMN table already contains predefined business domains for the Oracle client 2 Add mote records to the table by using a SQL script similar to the sample script in Figure 7 INSERT INTO KDD BUS DMN BUS DMN CD BUS DMN DESC TX BUS DMN DSPLY NM MANTAS DMN FL VALUES a Compliance Employees COMP N INSERT INTO KDD BUS DMN BUS DMN CD BUS DMN DESC TX BUS DMN DSPLY NM MANTAS DMN FL VALUES b Executives VEXECTC VN COMMIT Figure 7 Loading the KDD BUS DMN Table 3 Update the KDD_CENTRICITY table to reflect access to all focuses within the business domain with the following command update KDD CENTRICITY set bus dmn st a where KDD CENTRICITY CNTRY TYPE CD SC Refer to Creating Jurisdiction in tbe Database on page 35 to perform the
406. r which there is no location previously specified log syslog location Contains routing information for message libraries for this category for which there is no location previously specified log smtp hostname Identifies the hostname of the SMTP server if e mail address is specified as log output log fatal true Indicates that fatal logging is enabled false indicates that fatal logging is not enabled log fatal synchronous false Indicates that fatal logging is enabled false indicates that fatal logging is not enabled log warning true Indicates enabling of warning logging false indicates that warning logging is not enabled 308 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Logging Configuration File Appendix A Logging Table 118 Configurable Parameters for Common Logging Continued Property Sample Value Description log warning synchronous false Indicates enabling of warning logging false indicates that warning logging is not enabled log notice true Indicates enabling of notice logging false indicates that notice logging is not enabled log diagnostic log notice synchronous false Indicates enabling of notice logging false indicates that notice logging is not enabled false Indicates that diagnostic logging is not enabled true indicates enabling of diag nostic logging log diagnostic synchronous false Indicate
407. rading data Figure 16 runMDS sh Input and Output Directories Preprocessing When ingesting market data in flat files the Preprocessor can be used as an alternative Alternative tothe MDS to the MDS The following commands can be run in parallel lt INSTALL_DIR gt ingestion_manager scripts runDP sh InsideQuote INSTALL DIR ingestion manager scripts runDP sh MarketCenterQuote INSTALL DIR ingestion manager scripts runDP sh ReportedMarketSale The output of these commands will be the same as documented for the MDS The benefit of this approach is that the Preprocessor has some performance enhancements which the MDS does not If this alternative is used everywhere that this document refers to the MDS these three Preprocessors can be substituted Processing Data When the Ingestion Manager completes preprocessing of TC trading data and market through FDT and MDT data and prepared data output files the Firm Data Transformer FDT and Market Data Transformer MDT can rettieve them Upon completion of data preprocessing through scripts runDP sh and runMDS sh Ingestion Manager executes the runFDT sh and runMDT sh scripts The runMDT sh script can run as soon as runMDS sh processing completes However runMDS sh and runDP sh must complete before runFDT sh processing can begin 64 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Process Flow Chapter 4 Data Ingestion FDT Processing Duri
408. rann eee pde e en etie eedem pego ee De ee 229 Processing fof Ed DAleh eet to utet dum ertet e het idu e esveneailete eria ee iste ti prens 229 dentifyinga Running Batch Process uic enne pet P eR eerte pen 230 TT DR MOI TTA TEMPI 230 Batch Exp tt Utiliser iiuna as I ESRAS 230 OVELVIEW Ao id oed E i E E t A E EET My acdes oti 231 DECC E E E A TAAR A A E 231 Contiguration Parameters oeer aser tiri vites A NGAR 232 IBI sni iet EEA EE REEE RRR EERE EER RARER M 232 Script Details e rione e E REE EREE E E E Ea 233 AMtta CHMents E 233 XMI Metadata ean 233 Calendar Manager TOt iior tette to eb tet A ERE RAE R RE eiea 236 JOtteCtOty SCEUCDUEe eee einen eee e HOSP DEGERE jess GR HR GRE ANANO 236 DOGS E M N 236 Cal ndat Information EEUU OE 237 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide xi Contents Using the Calendar Manager Utility tret tenete toits eet epe sisisi 237 Conftouring the Calendar Manager URIy i e arre e tt reir ere a e te ES epar 237 Executing the Calendar Manager U Willy op El er epe E Pergit eene eed 238 Updating th KDD CAT TaD es s siters nen en RETRO nennt vtt do Yee 238 Data Retention Marnabetzicee ie onec ta RE det oreet R das dad Fas at e Cae ec 241 Direcoty o TUCU eT E 241 TOTS DRE 242 luces M 242 Usin
409. rchiving Process 113 Figure 26 Sample Fuzzy Name Matcher Utility install cfg File 121 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide xv List of Figures Figure 27 Figure 28 Figure 29 Figure 30 Figure 31 Figure 32 Figure 33 Figure 34 Figure 35 Figure 36 Figure 37 Figure 38 Figure 39 Figure 40 Figure 41 Figure 42 Figure 43 Figure 44 Figure 45 Figure 46 Figure 47 Figure 48 Figure 49 Figure 50 Figure 51 Figure 52 Figure 53 Figure 54 Figure 55 Figure 56 Figure 57 Figure 58 Figure 59 Figure 60 Figure 61 Figure 62 Figure 63 Figure 64 Figure 65 Figure 66 Sample install cfg Configuration Parameters sess 123 Optional Workflows AML Brokerage eee 136 Pre Watch List Workflows AML Brokerage eee 137 Watch List Workflows AML Brokerage eee 138 Post Watch List Workflows AML Brokerage eese 140 Summary Workflows AML Brokerage eee 143 Balances and Positions Workflows AML Brokerage oo eee 145 Optional Workflows AML Banking eeeeeeeennn 146 Independent Pre Watch List Workflows AML Banking 147 Watch List Workflows AML Banking eeeeeee 148 Post Watch List Workflows AML Banking see 150 Summary Workflows AML Banking eee 151 Optional
410. referential relationships within the data If an Oracle client chooses not to perform referential integrity checking grouping is not required except in some instances In this case a need still exists to process some reference data files prior to processing trading data These dependencies are as follows e Prior to executing the runMDS sh script you should ingest the following reference data files W Security m MarketCenter Prior to executing the runDP sh TradeExecution and runDL sh scripts you should ingest the following reference data files W Security m MarketCenter m CorporateAction m StructuredDeal m SettlementInstruction The ingestion process flow is as follows 1 Behavior Detection receives firm data in ASCII flat dat files which an Oracle client s data extraction process places in the inbox directory This data can be e Reference for example point in time customer and account data e Transactional for example market and trading data The preprocessor addresses only those files that match naming conventions that the DIS describes and which have the date and batch name portions of the file names that match the current data processing date and batch The Oracle client need only supply those file types that the solution sets require Ingestion Manager executes preprocessors simultaneously within hardware capacities The preprocessors use XML configuration files in the config datamaps directory
411. related party code 3 MLM Brokerage C 2192 w ph2192 update inst instn seq id ORION Common 2180 ommon Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 137 AML Brokerage Workflows Chapter 5 Informatica Workflows Informatica Watch List Workflows AML Brokerage Informatica Watch List workflows facilitate the application of customer supplied measures of risk to corresponding entities transactions and instructions Figure 30 illustrates Informatica Watch List workflows for AML Brokerage Wu 2 e Note Processing of workflows is not sequence dependent processing can occur simultaneously across all sequences lt Workflows with predecessors must wait for predecessor results before workflow Processing can begin Figure 30 Watch List Workflows AML Brokerage 138 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide AML Brokerage Workflows Chapter 5 Informatica Workflows The application completes these workflows in 12 sequences processing of some workflows requires data from a number of Miscellaneous predecessor workflows These workflows may proceed simultaneously in all sequences as soon as processing of any predecessors has completed Table 47 details the Watch List workflows in Figure 30 Table 47 Watch List Workflows AML Brokerage
412. ression job is part of the Behavior Detection subsystem Oracle Financial Services provides default job templates and job template groups for running Auto Close Alert You can modify these jobs using the Administration Tools Refer to the Oracle Financial Services Behavior Detection Platform Administration Tools User Guide Release 6 1 3 for more information To run the suppression job follow the steps 1 Verify that the dispatcher is running 2 Run the start_mantas sh script as follows start_mantas sh 507 where 507 is the job template that Oracle Financial Services provides to run the suppression job algorithm Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 183 Highlight Generation Chapter 6 Post Processing Tasks Highlight Generation The Alert Management subsystem displays alert and match highlights in the Alert List and Alert Context sections of the Oracle Financial Services UI The system calculates and stores these highlights in the database as part of the batch cycle using the following shell script The system generates highlights for alerts that appear in the alert list in the Alert Management subsystem and stores them to the database using the following shell script run_highlights ksh This script is part of the Database Tools that resides in the lt INSTALL_DIR gt database db_tools bin directory This script attaches to the database using the user that the utils data
413. restarting 220 using 216 utilities 215 alert suppression 183 Post Processing 183 AML Banking 150 151 AML Brokerage 140 150 analysis XML files null and padded space count 266 analysis xml file 263 constraints 264 distinct values 265 join counts 267 null and padded space count 266 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 327 Index annual activities KDD_CAL_HOLIDAY table 213 KDD_CAL_WKLY_OFF table 213 Loading Holidays 211 loading non business days 213 archiving cleanup activities 112 database 112 assignment 172 assignment job 177 auto close setting up rules 179 balances and positions 156 AML Brokerage 145 Broker Compliance 156 Batch Control Utility 8 195 223 configuring 225 directory structure 224 ending 229 identifying a running batch 230 intra day batch 226 KDD_PRCSNG_ BATCH_HIST table 227 KDD_PRCSNG_BATCH table 225 KDD_PRCSNG_BATCH_CONTROL table 227 logs 224 processing 227 setting up batches 225 single batch 226 starting 227 using 225 utilities 195 batch processing Administrative utilities 195 annual activities 195 Batch Control Utility 195 Calendar Manager Utility 195 Data Retention Manager 195 Database Statistics Management 195 ETL Process for Threshold Analyzer Utility 195 Flag Duplicate Alerts Utility 195 Process to Deactivate Expired Alert Suppression Rules 195 Push E mail Notification 195 Refreshing Te
414. rkflow truncates Account Summary Month Stage Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 323 Informatica Workflows Appendix D Informatica Workflow Details Table 120 Informatica Workflow Description Continued Workflow Number Workflow Name Description 0150 w p0150 truncate asms This workflow truncates Account Summary Month Stage 0180 w p0180 aggregate bot to file This workflow is used for the daily aggregation of the back office transaction data It populates the account summary file 0190 w p0190 aggregate fotps to file This workflow aggregates sums and counts various fields from the FOTPS for the account summary table It puts the results in a flat file for later processing 0200 w p0200 aggregate trade to file This workflow aggregates sums and counts various fields from daily trades for the account summary table It puts the results in a flat file for later processing 0270 w p0270 join retail acct activity This workflow joins retail account daily activity from files to asms and loads the data to account summary stage table 0310 w p0310 update asm for daily activ This workflow updates the Account Summary Month ity Table from its corresponding staging table 0470 w p0470 update asm for net worth This workflow updates Account Summary Month cur rent month data w
415. rkflows AML Banking e Informatica Miscellaneous Post Watch List Workflows AML Banking e Informatica Summary Wotkflows AML Banking Informatica Miscellaneous Optional Workflows AML Banking Optional Miscellaneous workflows perform processing in support of workflows in multiple functional areas Figure 34 illustrates Informatica Miscellaneous Optional wotkflows for AML Banking 0010 0070 Note Processing can occur simultaneously within each sequence Sequence 0 Workflows with predecessors 0080 0140 must wait for predecessor results before workflow processing can begin Figure 34 Optional Workflows AML Banking The application completes these workflows in one sequence processing of the wotkflows does not require data from external predecessors Table 51 describes the Informatica Miscellaneous Optional workflows in Figure 34 Table 51 Optional Workflows AML Banking Sequence Number Folder Workflow Number Workflow Name Predecessor 0 ORION Production 0010 w p0010 dump asm to file 0 ORION Production 0070 w p0070 reload asm from dump file 0 MLM Banking Pro 0080 w p0080 dump cbsm to file duction 0 MLM Banking Pro 0140 w p0140 reload cbsm from dump file duction 146 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide AML Banking Informatica Workflows Chapter 5 Informatica Workflows Informatica Miscellaneous
416. rkflows Fraud Detection 160 167 Web Application server 5 Web server 5 Where to Find More Information xxiv Who Should Read this Guide xxi 334 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide
417. rmation A historical data copy process copies alert related data that the Ingestion Manager provides from the source data tables to the historical data tables Refer to section Historical Data Copy on page 185 for more information An alert correlation process correlates alerts to business entities and optionally to each other based on configurable rule sets Refer to section Assignment on page 177 for more information Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 9 About Oracle Financial Services Operations Chapter 1 The Oracle Financial Services Behavior Detection Platform End Batch The system ends batch processing when processing of data from the Oracle client is complete Refer to section Ending a Batch Process on page 229 for more information The Alert amp Case Management subsystem then controls alert and case management processes Refer to the Oracle Financial Services Alert Management User Guide Release 6 1 3 for more information 10 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide CHAPTER 2 Behavior Detection Jobs This chapter provides an overview of the Oracle Financial Services Job Protocol and then explains how the System Administrator monitors jobs and starts and stops jobs when necessary In addition it describes the necessary scripts that you use for jobs This chapter focuses on the following topics About the Oracle
418. rnal predecessors These workflows may proceed simultaneously in all sequences as soon as processing of any predecessors has completed Table 58 describes the Summary workflows in Figure 41 The appearance of N A in sequence number and predecessor columns implies that the workflows are contingency workflows These workflows dump summaries to file and reload to the corresponding summary table Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 155 Broker Compliance Informatica Workflows Chapter 5 Informatica Workflows Table 58 Summary Workflows Broker Compliance Sequence Workflow Number Folder Number Workflow Name Predecessor 16 ORION Production 0150 w p0150 truncate asms ORION Common 2005 16 BSM Production 0171 w p0O171 truncate masms ORION Common 2005 16 BSM Production 0360 w p0360 aggregate block allcn day trades to file 16 BSM Production 0370 w p0370 aggregate block allcn trades to file 16 BSM Production 0380 w p0380 aggregate bot inter hh jrnls to file 16 9001 AccountDailyProfileTrade 16 9002 AccountDailyProfileTransaction 16 9003 InvestmentAdvisorProfile 16 9004 RegisteredRepresentativeProfile 17 ORION Production 0190 w p0190 aggregate fotps to asms ORION Common 3210 ORION Production 0150 17 BSM Production 0430 w p0430 join daily activity to masms BSM Production 0171 BSM_Pro
419. rocessing ORION Common 2180 2 ORION Common 2191 w ph2191 anticipatory profile instn processing ORION Common 2180 3 ORION Common 2193 w ph2193 INS instn processing ORION Common 2180 166 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Insurance Workflows Chapter 5 Informatica Workflows Informatica Watch List Workflows lnsurance Informatica Watch List workflows facilitate the application of customer supplied measures of risk to corresponding entities transactions and instructions Figure 50 illustrates Informatica Watch List workflows for Insurance External 3160 2190 2180 3120 3130 3091 Predecessor Misc Misc Misc Misc Misc Misc 3110 z 3171 Sequence 3 3005 3030 3000 Sequence 4 Sequence 5 Sequence 6 3180 Sequence 7 Sequence 8 Sequence 9 Sequence 11 Note Processing can occur simultaneously within each sequence Workflows with predecessors must wait for predecessor Sequence 12 results before Figure 50 Informatica Watch List Workflows Insurance The application completes these workflows in nine sequences processing of some workflows requires data from Miscellaneous predecessors These workflows may proceed simultaneously in all sequences as soon as processing of any predecessors has completed Table 67 details the Informatica Watch List workflows for Insurance in Figure 50 Oracle Financial Services Behavior Detection Platform 6 1 3
420. ronment for scenario migration modify the parameters that the Environment sample INSTALL DIR database db tools mantas cfg install cfg shows Refer to Table 107 on page 278 The tables in the following sections describe the parameters specific to the Scenario Migration Utility Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 277 Scenario Migration Utility Chapter 8 Administrative Utilities Configuring General Table 107 describes general scenario migration parameters Scenario Migration Table 107 General Scenario Migration Parameters Parameter Description score include Flag that indicates whether scenario migration includes scenario scoring metadata value is Y or N the default wrapper include Flag that indicates whether scenario migration includes wrapper metadata value is Y or N the default sm depfile Location of the scenario migration dependencies file INSTALL DIR gt data base db tools mantas cfg dep cfg sm release Version of the Scenario Migration Utility Caution Oracle Financial Services strongly recommends that you maintain scores and threshold values in a single environment Maintaining these attributes in multiple environments and migrating the scenarios between the environments can cause the loss of threshold set specific scoring rules 278 Oracle Financial Services Behavior Detection Platform 6 1 3 Administratio
421. rough the Report Framework s Report Management con sole Report KDD REPORT TRACK This table stores an instance of a This table is populated by report to be signed off the run report client ks h utility 314 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Report Management Appendix C Report Management Table 119 Metadata Creation for Report Framework Continued Logical Name Report Jurisdiction Table Name KDD REPORT TRACK JRSD CN Table Description This table stores the Jurisdictions associated to an instance of the report This table would be popu lated by the run report client kshultil ity Comments This table is populated by the run report client ks h utility Report Parameters KDD REPORT TRACK PARA M This table stores the parameter values used to filter the instance of the report This table is populated by the run report client ks h utility Report Comments Report Framework Limitations KDD REPORT TRACK CMMN T This table stores the comments that the user make against a report This table stores workflow or signoff comments added by the user Refer to the Oracle Financial Services Behavior Detection Platform FSDM Reference Guide Vol 1 Business Data for more information The report framework has the following limitations e Report Framework cannot import graphs and charts e Only the data
422. rs in the log file when skipping indexes XDL Default IndexParallelLev Specifies the parallel level of an index rebuild 4 el that is number of concurrent threads for rebuilding an index XDL Default DoAnalyze Specifies whether to run a stored procedure FALSE to analyze a database table after loading data into it XDL Default DoImportStatisti Specifies whether to run a stored procedure FALSE cs to import statistics for a database table after loading data into it XDL Default Specifies the type of statistic import to DLY POST LOAD ImportStatisticsType perform if DolmportStatistics has a value of True XDL Default Saves the directory to which the stored tmp ImportStatisticsLogDir procedure writes the log file if DolmportStatistics has a value of True This log directory must reside on the server that hosts the database XDL Default Specifies the database error code that 942 TableDoesNotExistErrorCode indicates a database table does not exist XDL Default UseUpdateLoader Specifies whether JDBC updates should be FALSE used instead of a delete insert when updating a database record This is only valid for data types that have a load operation of Update XDL Data Type gt lt Property gt Overrides the specified property for a given Data Loader instance IMC Specifies properties for configuring the Directory Cleanup IMC component Directory 1 N Name Id
423. rt The sm 1oad sh script loads translated XML table data files into the target database Correlation Rule To load alert correlation rule metadata follow these steps 1 Create a metadata configuration file someFileName cfg with the rule names of the alert correlation rules to be loaded corr rule name in the lt someFileName gt cfg file and or corr rule name corr rule file name gt Note Providing both corr rule name and corr rule file name in the lt someFileName gt cfg file allows the user a flexibility to specify the actual filename that contains the metadata information for the respective alert correlation rule If you provide both an alert correlation rule name and an alert correlation rule file name on a line you must separate them with an equals sign It is recommended that the alert correlation rule file name be specified without an extension 2 Navigate to the following directory cd INSTALL DIR gt database db tools mantas cfg 3 Edit the install cfg file to include the path for the above created file against the tag corrRuleMig CorrRuleFileNm 4 Copy the XML files you plan to load into the directory that the load dirname specifies in the install cfg file 5 Navigate to the following directory cd INSTALL DIR gt database db tools bin 6 Execute the sm 1oad sh script as follows sm load sh correlation Note The utility performs the following validations upon loading of an
424. rvices Behavior Detection data ingestion process completes refer to Chapter 4 Data Ingestion on page 51 for more information This chapter focuses on the following topics e About Informatica Workflows e AML Brokerage Workflows e AML Banking Informatica Workflows e Broker Compliance Informatica Workflows e Fraud Detection Informatica Workflows e Insurance Workflows About Informatica Workflows After the Oracle Finanical Services Behavior Detection Ingestion Manager loads the base tables the process of deriving and aggregating data begins The client s job scheduling system invokes Informatica workflows that perform this data manipulation It is the client s responsibility in consultation with Oracle Financial Services technical staff to configure the job scheduling system for successful completion of this processing Processing uses the lt Informatica install directory gt Scripts start_wkflw sh script to invoke individual workflows The command includes the location and name for each workflow type that requires processing in the following format Informatica install directory gt Sctipts statt_wkflw sh folder lt workflow name gt For example Informatica install directory Scripts start wkflw sh MLM Brokerage Common w ph3041 create addresses from inst Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 133 About Informatica Workflows Chapter 5 Informatica Workflows
425. s sss 11 About the Oracle Financial Services Job Protocol essent tentent tene tenentes 11 Understanding the Oracle Financial Services Job Protocol sss 12 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide v Contents Understanding the Dispatcher Process ivsessiisssssssscssssssscsscctsosssanosersevenssvanssvenssvatiensvstssssiessetessstesvsboosssesssssosetussetsetees 13 Understanding the maritas Process sii iis ssisss sissssessisoelossisondaisscogessdessseasanseiusedbasiedssnsesedsdotvivendovsieosinisiervivendovedessiededers 13 Applying a Dataset Ovettide s ee Reip E RP I ERR ER E A a E a a 13 Configuring the Dataset Override Peature i dst tert eight tae tte reete et eit eden 13 Pertorminp Dispatcher Tasks ase at tus P RETURN bet e ue er EARTH tr a e eiue 15 Setting Environment Variables C 15 About DDO system env Fue secs sas seta da tee ttal a e ted ee dasa 15 Starting the BUT elo a 16 Stopping the Dispatchets icto tee eter e oue opes en ever e EOS ERE SER I 17 Monitoring the Dispatchet i nace ir ne d a a d p eR ese cA R 17 Performing Job Tasks3 ioco o eter re Pe PRIN CN PR IR desee UR CAR Vet ee apto FIR PNEU AERA 18 Understanding the Job Status Codes tette tiet itii rires rites testet to into es ire tod a e 18 Startins M
426. s The log file contains relevant information such as status of the purge processing log relevant information and error records You can modify the current logging configuration for the Alert Purge Utility in the INSTALL DIR database db tools mantas cfg install cfg and categories cfg files For more information about logging in these configuration files Refer to Common Resources for Administrative Utilities on page 198 and Appendix A Logging on page 299 for more information Precautions You use the utility to rid the system of falsely generated matches and alerts Other than recorded information in the INSTALL DIR database db tools logs Alert Purge log file the system does not capture audit information for this process The utility does not update other alerts prior counts as a result of purging alerts Note You cannot purge an alert that is used to trigger Auto Suppression You can tell if an alert ID is used to trigger Auto Suppression by looking at the kdd auto suppr alert trgr alert id column to see if it contains the alert ID in question If so you have to delete the record before attempting to purge the alert Run the Alert Purge Utility e Through one process at a time Multiple simultaneous executions of the utility may lead to unexpected results and compromise the relational integrity of match alert and action data e When no users are editing or viewing any of the alerts actions or associated infor
427. s Behavior Detection Platform 6 1 3 Administration Guide 97 Data Ingestion Directory Structure Chapter 4 Data Ingestion Table 32 Datalngest xml File Configuration Parameters Continued Property Name Description Example e X x DP Default SequenceBatchSiz DP Default AdditionalOutput Specifies the batch size when retrieving sequence IDs Specifies a directory to contain the output file in addition to the target directory 100000 DP Default DoFileLookups Specifies whether to do reference data lookups for fields that arrive as part of an input file TRUE or not do them FALSE FALSE Pas lures DP Default DiscardLookupFai Specifies whether to discard records that fail a reference data lookup TRUE or just log a message FALSE FALSE DP Default ValidatorClass Specifies the Java class used to validate records of a given data type Use of subclasses occurs when the general functionality of AbstractValidator is not enough for a given data type AbstractValidator XDP Default AdapterClass Specifies the Java class used to process records of a given data type Use of subclasses occurs when the general functionality of BaseFileAdapter is not enough for a given data type BaseFileAdapter XDP Default NumberOfThreads Specifies the number of worker threads to be used when preprocessing a file XDP XDP Default BufferSize Def
428. s of business data partitions and market data partitions differ somewhat e Business data partitions are predefined so that weekdays Monday through Friday are business days and Saturday and Sunday are weekly off days Business data tables use all partitioning types Howevet you can use the Calendar Manager Utility to configure a business calendar as desired For more information about this utility Refer to Calendar Manager Utility on page 236 for more information e Market data partitions hold a single day of data The partitions use the PYYYYMMDD convention where YYYYMMDD is the date of the partition You should run partition maintenance as appropriate for your solution set Oracle Financial Services recommends that you run partition maintenance for AML on a daily basis after setting the business date through the Calendar Manager Utility and ptior to the daily execution of batch processing and Trading Compliance at least once a week Note Oracle Financial Services recommends that you use the P Partition option when running the Data Retention Manager as it drops older partitions and adds appropriate partitions in a single run of the utility When performing monthly maintenance you can add or drop a partition independently as the following procedures describe As part of an alternative method of monthly partition maintenance you can either add or drop a monthly database partition as the following sections describe
429. s of the various Calendar Manager processes results and error records You can modify the current logging configuration for this utility in the configuration files INSTALL DIR database db tools mantas cfg install cfg and categories cfg For more information about logging in these configuration files Refer to Common Resources for Administrative Utilities on page 198 and Appendix A Logging on page 299 for more information 236 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Calendar Manager Utility Chapter 7 Batch Processing Utilities Calendar Information The Calendar Manager Utility obtains all holidays and weekly off days for loading into the calendars by retrieving information from the KDD_CAL_HOLIDAY and KDD CAL WKLY OFF tables Refer to Table 77 and Table 78 These tables contain calendar information that an Oracle Financial Services client has provided regarding observed holidays and non business days Using the Calendar Manager Utility Configuring the Calendar Manager Utility The Calendar Manager Utility runs as part of automated processing that a job scheduling tool such as Maestro or Unicenter AutoSys controls The utility runs through a shell script using values in parameters that particular configuration files contain The utility then populates the KDD CAL database table with relevant business calendar information The following sections desctibe this process includ
430. s summaties of trading transaction and instruction activity e Assigns transaction and entity risk through watch list processing Updates various Balances and Positions derived attributes Note To successfully run Informatica workflows you must have installed Informatica Refer to the Oracle Financial Services Behavior Detection Platform Installation Guide for more information The system uses Informatica to perform these tasks Figure 20 illustrates Informatica processing Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 69 Process Flow Chapter 4 Data Ingestion 1 Reads 2 Applies mappings from mappings to data repository in FSDM 3 Results in derivations and aggregations Example FSDM contains account activity for a particular day in the form of trades transactions and or instructions Informatica applies appropriate mappings from its repository and updates the account for daily activity Figure 20 Informatica Workflow Processing Informatica does the following e Reads mappings or workflows in its repository e Applies relevant workflows to FSDM Reference Transaction and Derived data e Updates summary tables Figure 21 Watch List content and various Balance and Positions derived attributes Figure 22 70 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Process Flow Chapter 4 Data Ingestion Figure 21 Inf
431. s that diagnostic logging is not enabled true indicates that diagnostic log ging is enabled log trace false Indicates that trace logging is not enabled true indicates enabling of trace logging log syslog hostname log trace synchronous true Indicates that trace logging is not enabled true indicates enabling of trace logging hostname Indicates the host name of syslog for mes sages sent to syslog log smtp hostname hostname Indicates the host name of the SMTP server for messages that processing sends to an e mail address log time zone EST Indicates the time zone that is used when logging messages Monitoring Log Files The Ingestion Manager uses common logging by assigning every component for example FDT or MDT a category You can configure the destination of log messages for each component which Table 117 describes The default logging behavior is to send log messages to the component s designated log file in the date subdirectory representing the current processing date under the logs directory This behavior can be turned off by setting the LogeUseDefaultLog attribute in DataIngest xml to false If this attribute is true the system continues to send messages to the designated log files in addition to any behavior that the common logging configuration file specifies When using a tool to monitor a log file use the message ID to search for a particular log message instead of
432. s the conventions used in this guide Table 1 Conventions Used in this Guide This convention Stands for Italics Names of books chapters and sections as references Emphasis Bold Object of an action menu names field names options button names in a step by step procedure Commands typed at a prompt User input Monospace Directories and subdirectories File names and extensions Process names Code sample including keywords and variables within text and as separate paragraphs and user defined pro gram elements within text lt Variable gt Substitute input value Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide XXV Conventions Used in this Guide About This Guide xxvi Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide CHAPTER 1 The Oracle Financial Services Behavior Detection Platform This chapter provides a brief overview of the Oracle Financial Services Behavior Detection Platform in terms of its architecture and operations It also includes new features for this release This chapter focuses on the following topics e Technology Compatibility About the Oracle Financial Services Architecture e About Oracle Financial Services Operations Technology Compatibility Oracle Financial Services is able to meet the environmental needs of its customers by providing support for third party tools
433. s the logic to identify trades which require additional analysis This procedure updates the REQ ANALYSIS FL column of the KDD TRADE BLOTTER table setting it to Y for trades requiring additional analysis To augment trade blotter data run the following command run augment trade blotter sh yyyymmdd where yyyymmdd is an optional input parameter If batch date lt yyyymmdd gt is not provided the system takes the 184 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Score Trade Blotter Chapter 6 Post Processing Tasks current batch date from the DATA_DUMP_DT column of the KDD PRCSNG BATCH CONTROL table The log for this script is written in the run stored procedure 1log file under the lt INSTALL_DIR gt database db_tools logs directory This script is a part of the database tools and resides in the INSTALL DIR database db tools bin directory Note This utility can be run anytime after data ingestion of Trade Blotter has been successfully completed Score Trade Blotter There is certain information that must be processed in order for the Alert Management system to be able to display the Trade Blotter data This includes the score of the trades and the mapping between alerts and trades The system can determine the maximum score of alerts generated in the same batch cycle associated with a trade as well as determine the alert trade mappings by the execution of the follow
434. s to Standard Data Ingestion Practices Table 26 describes alternatives to the following Data Ingestion processing options and provides advantages and disadvantages to each option e T 1 vs intra day e Single source vs multiple sources e Live market data vs file based market data e Truncating reference data vs updating reference data e Single instance vs multiple instances Table 26 Data Ingestion Options and Alternatives Typical Process Alternative Process T 1 Process by which the current day s data becomes available on the following day or at the end of the current day Advantages e Simplifies batch processing and scheduling e Supports truncating and reloading of reference data e Provides an ideal mechanism for executing AML and many TC scenarios e Provides support for multiple batch processing e Processing of data originates from one source in a single batch Disadvantages e Delays availability of alerts until the following day e Limits the batch processing time window especially for multi regional data Intra day Process by which Data Ingestion occurs on the day that the data becomes available Intra day processing focuses on collecting transforming and loading all Market and Firm data during market trading hours and throughout non trading hours if necessary This allows viewing of detection results on the same day as generation of data Advantages Use of data in detec
435. s who fall under that pool If the user who created the case is also assigned as the owner no notification is generated e Re assigned case notification Notification is generated to new owner of the case upon reassignment of the case If the user who reassigned the case is also the new ownet no notification is generated If the new owner is a pool then notification is generated to all users who are members of the organization represented by that pool e Re assigned alerts notification Notification is generated to the new owner of the Alert upon reassignment of the alert If the user who reassigned the alert is also the new owner no notification is generated If the new owner is a pool then notification is generated to all users who are members of the organization represented by that pool e Alert Data Transfer Unsuccessful In Asynchronous alert data transfer mode if there is Unsuccessful data transfer during promotion of an alert to a case or linking of an alert to a case then notification is generated to the User who is taking the action the owner of the alert the owner of the case and the assigned to uset of the case Batch Based These notification are result of processing of end mantas batch sh Following are the batch based notifications e Cases Near Due Date notification Notification is generated to the owner of the cases if the due date of the case falls within the configurable parameter set in the Installation parameter tabl
436. schema which the schema business owner property identifies The script logs into the database using the user that the truncate database username property specifies in the INSTALL DIR database db tools mantas cfg install cfg file The script has the following calling signature run truncate manager sh table name gt Note This process is not intended to be called independently only the Ingestion Manager subsystem should use it 258 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide ETL Process for Threshold Analyzer Utility Chapter 7 Batch Processing Utilities ETL Process for Threshold Analyzer Utility For inserting and updating records into the KDD_TA_ML_DATA KDD TA BC DATA and KDD TA TC DATA tables there are two shell scripts that are used to call the database procedures These are run insert ta utility sh This script calls the P TA ML INSERT BREAKS P TA BC INSERT BREAKS and P TA TC INSERT BREAKS procedures which insert data into the KDD TA ML DATA KDD TA BC DATA and KDD TA TC DATA tables respectively based on the CREAT_TS of the alerts in relation to the LAST RUN DT from KDD TA LAST RUN values for RUN TYPE CD ate ML I BC I and TC I There is one optional parameter DEBUG_FL for this shell script defaults value to FALSE If you provide a value of TRUE as an argument then information insert commands is also loaded into the KDD TA INS DEBUG table It also upd
437. seeeeeenee tentent tenerent 187 BUSINESS Enny PADS iss ose sesen toi Eo toed snes teak e teet o n a oi d adus 187 Correlation Rule netter RE EROR Sut esata caveats cove sauces Hs irati A prede 188 Activating or Deactivating Correlation Raes eee ee e e tenente tentent tenent tente E ete nen rene 191 dample Alert Correlation IS eicit t E en eit PEDE P eT ed i era 192 Account Approval Pre Trade Approval TASkS aaseunsue dne Sepe M dta duum n Rn d odd equ 192 x Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Contents CHAPTER 7 Batch Processing Utilities ssssssssssssssse 195 About Administrative Utilities itte tont tenete stehe tette atero soos eee von iere oa E E EERE 195 Common Resources for Administrative Utilities sse tette 198 mpi M RRR 198 DAI HaT EE 208 Configuring NOT M 210 About Annual A CUVINES RC M 211 Loading Holidays atto V ba e aud d AR Pb nodi editore keta 211 I oadinie Nonzbusiness DAys or pee eere epe thee teet I het pere e epit be pU sxe 213 Poder nnani arna aa E E a Eia A E E E E A ai 215 Directory Structuren eee eE EEEE EENEN EENE AEE EE EEEE EEEE EENEN CEEE EEEE EEEE EEEE 215 DOE Sinse e OE REENE PREE E O OT E EEE EAA EESEL 216 Pr cautiOhs inen dee tete A S E AE RE E sud radios 216 Usin the Alert Puree Unli
438. sequence Workflows with predecessors must wait for predecessor results before workflow processing can begin Figure 28 Optional Workflows AML Brokerage The application completes these workflows in one sequence processing of the workflows does not require data from external predecessors Table 45 describes the Optional Miscellaneous workflows in Figure 28 Table 45 Optional Workflows AML Brokerage Sequence Number Folder Workflow Number Workflow Name Predecessor 0 ORION Production 0010 w p0010 dump asm to file 0 MLM Brokerage P 0020 w p0020 dump csm to file roduction 0 MLM Brokerage P 0030 w p0030 dump iasm to file roduction 0 ORION_Production 0070 w_p0070_reload_asm_from_dump_file 0 MLM Brokerage P 0090 w p0090 reload csm from dump file roduction 0 MLM Brokerage P 0100 w p0100 reload iasm from dump file roduction 0 MLM Brokerage P 2020 w p2020 delete nonstanding instructions roduction 136 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide AML Brokerage Workflows Chapter 5 Informatica Workflows Informatica Miscellaneous Pre Watch List Workflows AML Brokerage Pre Watch List Miscellaneous workflows perform processing in support of workflows in multiple functional areas Figure 29 illustrates Pre Watch List Miscellaneous Informatica workflows for AML Brokerage Sequence 1 2005 runNUAM
439. ser authentication and authorization It also contains instructions for setting up user accounts in the database to access the Scenario Manager This chapter focuses on the following topics About the Oracle Financial Services User Authentication Setting up a User About Configuring Access Control Metadata Mapping Users To Access Control Metadata About Scenario Manager Login Accounts About Changing Passwords for System Accounts About Configuring File Type Extensions About the Oracle Financial Services User Authentication The primary way to access information is through a Web browser that accesses the Alert Management Case Management Account Approval Pre Trade Approval and Administration Tools The application offers SMS for authentication of web browser clients The Scenario Manager authenticates use of the Behavior Detection database only Behavior Detection offers two authentication mechanisms for Web browser clients Built in Authentication System and Security Management System SMS on the Web Application server that authenticates users from a login Web page Refer to section Understanding SMS on page 26 for more information Web server authentication for Oracle clients who want to utilize their own External Authentication Management EAM tool Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 25 About the Oracle Financial Services User Authentication Chapter 3 Security Confi
440. sh Sequence 2 Note Processing can occur within Sequence 3 each sequence simultaneously Workflows with predecessors must wait for predecessor results before workflow processing can begin Figure 29 Pre Watch List Workflows AML Brokerage The application completes these workflows in three sequences processing of the wotkflows does not require data from external predecessors Table 46 describes the Informatica Miscellaneous Pre Watch List workflows in Figure 29 Table 46 Pre Watch List Workflows AML Brokerage Sequence Workflow Number Folder Number Workflow Name Predecessor 1 ORION Common 2005 w p2005 set cd pd dates 1 MLM Brokerage P 2040 w p2040 update cust count of accounts roduction 1 ORION Common 2050 w ph2050 update bot reversals 1 ORION Common 2070 w ph2070 truncate cas 1 MLM Brokerage C 2071 w ph2071 truncate das ommon 1 ORION Common 2140 w ph2140 pass thru process 1 ORION Common 2180 w ph2180 instn identification 1 ORION Common 2450 w ph2450 populate ACCT SRVC TEAM 1 ORION Common 2600 w ph2600 loan smry mnth 2 MLM Brokerage C 2122 w ph2122 aggregate tsv offsetting trades to tdtcs MLM Brokerage Common 2071 ommon 2 ORION Common 2190 w ph2190 fotps instn processing ORION Common 2180 2 ORION Common 2191 w ph2191 anticipatory profile instn processing ORION_Common 2180 3 ORION Common 2130 w ph2130 update fot un
441. shes removing old data subdirectories and their contents env sh Contains common configuration settings required to run Data Ingestion subsystem components The run sh and stop sh scripts use this script truncate table sh lt schema tablename gt runUtility sh Truncates a specified table in the database Processing runs this script prior to loading reference data when an Oracle client wants to perform a full refresh of the data Launches a Java based utility to derive the contents of a given database table You datatype need to run runDL sh data type after this script has executed successfully For example runUtility sh AccountProfile runDL sh AccountProfile 84 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Ingestion Directory Structure Chapter 4 Data Ingestion config Subdirectory The run scripts in Table 28 configure the executing environment for the Java component and then execute it All run scripts invoke the env sh script to define environment variables that the components require The run scripts also start the Java program with appropriate command line parameters which Table 29 describes Table 29 Environment Variable Descriptions Parameter Description classpath Directs the Java Runtime Environment JRE to the location of Java programs and supporting Java classes Djava security polic ye Sets the location of the polic
442. specified logging displays on the console log default location mantaslog Specify the location directory path of the mantaslog if the mantaslog was chosen as the log output location anywhere above Logging displays on the console if mantaslog was selected and this property is not given a value log mantaslog location tmp Specify the hostname of the SMTP server if an e mail address was chosen as the log output location anywhere above Logging will go to the console if an e mail address was selected and this property is not given a value log smtp hostname 122 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Fuzzy Name Matcher Utility Chapter 4 Data Ingestion Using the Fuzzy Name Matcher Utility Configuring the Fuzzy Name Matcher Utility The utility typically runs as part of automated processing that a job scheduling tool such as Maestro or Unicenter AutoSys manages You can also execute the utility through a UNIX shell script which the next section describes The following topics describe this process e Configuring the Fuzzy Name Matcher Utility e Executing the Fuzzy Name Matcher Utility In addition to log configuration parameters the software informatica mantas deployment ingestion_manager fuzzy_match install cfg configuration file contains the configuration information that the Fuzzy Name Matcher Utility requires for processing Figure 27
443. ss 0 5 Expand All Remove Case Type Subtype P Case Type Subtype 0 3 Expand All Remove Correlation Rule LEM Correlation Rule 0 Remove User iv USER TestOrgA Yes Yes pavan oracle com Choose User GEN INST Business Domain Name GEN INST Help p in Business Domain Description General Institutional Broker Dealer BI Figure 10 Components of Security Attribute 2 Local intranet fa 100 Note In order to update the user profiles before proceeding with mapping any security attributes select the value User from the Choose User Type drop down list When chosen all the updates made to user profiles through the User Maintenance UI are imported from the CSSMS_USER_PROFILE table of the OFSAAI configuration schema to the KDD_REVIEW_OWNER table of the mantas schema This action does not affect the security attributes that might be already mapped Once the user details are imported the security attributes should be mapped remapped The drop down lists have options for both Organizations and Users To map an organization select the organization from the drop down list and select the corresponding Organization in the Choose User drop down list Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 43 Mapping Users To Access Control Metadata Chapter
444. ss gt Note that category names that is property values cannot contain the following reserved words fatal warning notice diagnostic trace category or location You can configure multiple locations for each property using a comma delimiter For example log category TEST CATEGORY location console mantaslog log category TEST CATEGORY 2 location console users jsmith logs mylog log If an Oracle Financial Services client chooses to record log messages to files those log files may become very large over the course of time or depending on the types of logging enabled If this occurs the system rolls files larger than 10 MB if log max size property is not specified over to another log file and adds a number incrementally to the log file name For example if your log file name is mantas 109g additional log files appear as mantas 10g 1 mantas 1og 2 so forth Note The maximum value for the 1og max size property can be 2000000000 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 307 Logging Configuration File Appendix A Logging Configurable Logging Properties Table 118 identifies the configurable properties for logging in an Oracle Financial Services client s environment Table 118 Configurable Parameters for Common Logging Property Sample Value Description log format d t Yop Yom n Identifies the log formatting string Refer to Apache
445. ssion for a shortened day The format is HHMM SESSN CLS TM Indicates the closing time of the trading session for a shortened day The format is HHMM SESSN TM OFFST TX Indicates the timezone offset for SESSN OPN TM and SESSN CLS TM The format is HH MM Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 239 Calendar Manager Utility Chapter 7 Batch Processing Utilities Table 93 KDD_CAL Table Contents Continued Column Name Description WK_BNDRY_CD Week s start day SD and end day ED e If this is the last business day for this calendar name for the week that is next business day has a lower DAY OF WK value set to ED lt x gt where x is a numeric counter with the start end of the week that the provided date is in 0 e If itis the first business day for this calendar name for this week that is previous business day has a higher DAY OF WK value set to SD lt x gt Weeks before the provided date increment the counter and weeks after the provided date decrement the coun ter Therefore EDO is always on the provided date or in the future and SD0 is always on the provided date or in the past MNTH BNDRY CD Month s start day SD and end day ED e lf this is the last business day for this calendar name for the month that is next business day in a different month set to ED lt y gt where y is a numeric counter with the s
446. st place your query within a CDATA block e Verify that no white space exists between the SOL query opening tag and the CDATA tags for example CDATA and the closing tag for example tellak 270 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Analysis Tool Chapter 8 Administrative Utilities e Processing extracts column headers in the output from the SQL query itself When performing calculations in return columns it is best to alias the return columns for output e Line breaks and comments in the SQL are acceptable but you should use style comments in lieu of single line comments for safety e The tool does not perform any schema name substitution Therefore verify that any schema names match the database contents The database user for example INGEST USER has aliases for most tables you may need to analyze Thus running the tool as INGEST USER should prevent you from needing schema names in queties Using the Data Analysis Tool After editing the configuration files you can run the Data Analysis Tool as a foreground or background process Table 102 lists the XML input files delivered for use with the Data Analysis Tool Table 102 Data Analysis Tool XML Input Files File Description analysis aml xml Analysis configuration specific for data required by Anti Money Laundering scenarios and Ingestion Manager operations to support them analysis aml ui
447. stion manager scripts runDP sh Account Ingestion Manager processes data files in groups in a specified order from Oracle client data in the inbox directory Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 57 Process Flow Chapter 4 Data Ingestion Table 19 lists the data files by group Table 19 Data Files by Group AccountCustomerRole AccountPhone AccountEmailAddress AccountRealizedProfitAndLoss Country EmployeeTolnsurancePolicy FrontOfficeTransaction InsuranceProduct InsurancePolicy InsurancePolicyBalance InsuranceSeller InsuranceSellerToLicense InsuranceTransaction Issuer InsurancePolicyToCustomer InsurancePolicyFeature CollateralValueCurrency SecuritiesLicense ExternallnvestmentAccountPosition AccountGroup SecurityFirmDaily TrustedPair AccountAverageNetWorth FirmAccountPositionPair NaturalGasFlow PeerGroup Account Customer Employee FrontOfficeTransactionParty MarketTradingSession OrganizationRelationship AccountGroupAddress AccountGrouplnvestmentObjective AccountGrouplOSMember AccountGroupMemberExperience BankerToOfficer GeneralUsageList LoanOriginationAction InvestmentGuideline InvestmentGuideline ToAccount Letteroflntent Loan LoanDailyActivity MarketCenter MarketIndex MarketIndexDaily Organization RegisteredRepresentativeComplaint ServiceTeamMember SystemLogonType WatchList OnlineAccount CollateralValueProduct EnergyAndCommodityInstrument
448. stration Guide 125 Fuzzy Name Matcher Utility Chapter 4 Data Ingestion Executing the Fuzzy Name Matcher Utility To Execute the Fuzzy Name Matcher Utility To execute the Fuzzy Name Matcher Utility manually use the following procedure To execute the Fuzzy Name Matcher utility type the following at the UNIX command line fuzzy match sh t target name list c candidate name list r result file where target name list Name of the target name list file names for which you are searching This is a delimited file that contains the following tokens on each line m Target name m Delimiter which is defined in the install cfg file m Business B or Personal P name candidate name list Name of the candidate name list file names among which you are searching This file contains one candidate name per line result file Name of the results file Note The utility uses only the following characters in matching a z hyphen apostrophe 0 9 It considers any other character as a separator between tokens Also the target names and candidate names must not contain delimiter characters If the utility completes successfully it returns 0 at the command line if it does not it returns 7 Using Informatica Identity Resolution API for Name Matching The Fuzzy Name Matcher utility is used by two processes Informatica workflows and Scan Watch List Web Service within Behavior Detection In Infor
449. subject to change in subsequent releases Folder Folder location of each workflow Workflow Number Unique four digit number that represents a particular work flow Workflow Name Unique name of each workflow AML Brokerage Workflows The following sections describe the Informatica workflows that are required for deriving and aggregating data for the AML Brokerage solution e Informatica Miscellaneous Optional Workflows AML Brokerage e Informatica Miscellaneous Pre Watch List Workflows AML Brokerage e Informatica Watch List Workflows AML Brokerage e Informatica Miscellaneous Post Watch List Workflows AML Brokerage e Informatica Summary Workflows AML Brokerage e Informatica Balances and Positions Workflows AML Brokerage Each section provides a graphic that illustrates the workflow An accompanying table desctibes the process by sequence workflow number and name and internal or external predecessors if any Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 135 AML Brokerage Workflows Chapter 5 Informatica Workflows Informatica Miscellaneous Optional Workflows AML Brokerage Optional Miscellaneous workflows perform processing in support of workflows in multiple functional areas Figure 28 illustrates Optional Miscellaneous Informatica workflows for AML Brokerage Sequence 0 Note Processing can occur simultaneously within each
450. t dataset customizations specific to your site which can be retained outside of the scenario metadata The override to a dataset definition is stored in a file accessible by the Behavior Detection engine The dataset overtide feature allows improved performance tuning and the ability to add filters that are applicable only to your site s dataset When the system runs a job it retrieves the dataset definition from the database The Behavior Detection engine looks in the configured directory to locate the defined dataset override The engine uses the override copy of the dataset instead of the copy stored in the scenario definition in the database if a dataset override is specified The following constraints apply to overriding a dataset e The columns returned by the dataset override must be identical to those returned by the product dataset Therefore the dataset override does not support returning different columns for a pattern customization to use e The dataset override can use fewer thresholds than the product dataset but cannot have more thresholds than the product dataset Only thresholds applied in the dataset from the scenario are applied If a dataset override is present for a particular dataset the override applies to all jobs that use the dataset The following section provides instructions to configure the directory for the Behavior Detection engine for locating the defined dataset override To configure a dataset override
451. ta loaders 67 Watch List workflows 160 167 data rejection 112 loading 112 preprocessing 109 transformation 110 Data Retention Manager 195 241 configuring 244 creating partitions 247 directory structure 241 executing 245 KDD_DR_JOB table 250 KDD_DR_MAINT_OPRTN table 249 KDD_DR_RUN table 251 logs 242 maintaining partitions 247 processing 242 work tables 249 database partitions 247 creating 247 indexes 249 maintaining 247 recommended maintenance 248 Database Statistics Management 251 KDD_ANALYZE_PARAM table 252 logs 251 using 251 utilities 196 dataset override 13 using 13 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 329 Index Delta Mode 62 deployment view 5 Detection Algorithms C behavior detection algorithms 4 directory structure 215 Alert Purge Utility 215 Calendar Manager Utility 236 subdirectories 80 Dispatcher Process 11 stopping 17 dispatcher process 13 monitoring 17 shell script 12 starting 16 tasks 15 EAM 25 Environmental Variables 15 execute Calendar Manager Utility 238 External Authentication Management 5 Extract Workflow 52 F Financial Services Data Model 4 Firm Data Transformer FDT 64 processing 65 Flag Duplicate Alerts Utility 253 executing 253 Flag Duplicate Alerts utilities 196 using 253 Fraud Detection 163 FSDM 69 Full Refresh 62 Fuzzy Name Matcher Utility 120 126 confi
452. table sh table name gt view name where table name identifies the name of the table to populate view name identifies the name of the view to run to populate the table This procedure deletes all records in the target table prior to running the view to populate it It then estimates statistics for the newly populated table This procedure logs into the database with the user that the utils miner user property identifies in the INSTALL DIR database db tools mantas cfg install cfg file Populate Temporary Tables for Scenarios IML HiddenRelationsh ips dINST Scenarios typically depend on Data Ingestion to complete processing however the IML HiddenRelationships dINST ML NetworkOfAcEn fAC and CST LOSSES scenarios depend on population of Temp Tables to populate data The Link Analysis scenatio also depends on the network job creation before the sequence matcher part of the scenario runs To populate the temporary tables for IML HiddenRelationships dINST scenario follow the steps 1 Execute these refresh temporary table processes these commands can be run in parallel DB TOOLS refresh temp table sh TMP HIDREL NT JRNL TMP HIDREL NT JRNL VW Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 255 Refreshing Temporary Tables Chapter 7 Batch Processing Utilities ML NetworkOfAcEn fA C is 114698616 MP HIDREL NT CUACEMAIL VW
453. tage and offers guidance on ways to resolve rejections refer to section Rejection During the Preprocessing Stage on page 109 for more information e Rejection During Transformation Stage describes how rejections occur during the Transformation stage and offers guidance on ways to resolve rejections refer to section Rejection During the Transformation Stage on page 110 for mote information e Rejection During Loading Stage describes how rejections occur during the Loading stage and offers guidance on ways to resolve rejections refer to section Rejection During the Loading Stage on page 112 for more information Rejection During the Preprocessing Stage The first stage of ingestion is Preprocessing At this stage Data Ingestion examines Oracle Financial Services client reference and trading data for data quality and format to ensure the records conform to the requirements in the DIS Common reasons for rejection of data during Preprocessing include problems with data type missing data referential integrity and domain values During normal operation the number of rejections at the preprocessor stage should be minimal If the volume of rejections at this stage is high a decision threshold can halt processing and allow manual inspection of the data The rejections are likely the result of a problem in the data extraction process It is possible to correct the rejections and then reingest the data Oracle Financial Services Beh
454. tains a symbol s InsideQuote MarketCenterQuote and ReportedMarketSale information The IMC component runIMC sh determines the length of time that a Market data file remains in the subdirectory before its removal The IMC s configuration file defines the number of days that market data files persist before removal The transform subdirectory within the market subdirectory contains the MDT s checkpoint data and working files that it creates during transformation The MDT receives preprocessed data that MDS creates and transforms that data to create derived attributes Processing writes the transformed data to files and moves the files to the load subdirectory upon completion The MDT also maintains checkpoint files that allow it to recover after a failure and without the loss of data integrity the MDT removes the files after it transforms its data successfully Table 34 identifies the files that the MDT writes to subdirectories within the Load subdirectory Table 34 Files Output by Market Data Transformer Component Output Data Files MDT InsideQuote MDT MarketCenterQuote MDT ReportedMarketSale MDT 104 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Ingestion Directory Structure Chapter 4 Data Ingestion load Subdirectory data backup Subdirectory data firm Subdirectory extract Subdirectory The load subdirectory within the market subdirectory co
455. tart end of the month that the provided date is in 7 O e f itis the first business day for this calendar for this month that is previous business day in a different month set to SD lt y gt Months before the provided date increment the counter and months after the provided date decrement the coun ter Therefore EDO is always on the provided date or in the future and SDO is always on the provided date or in the past If a batch is running the system uses the date provided in the call to start the set mantas date sh script This script updates the KDD PRSCNG BATCH CONTROL DATA DUMP DT field 240 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Retention Manager Chapter 7 Batch Processing Utilities Data Retention Manager Directory Structure Behavior Detection relies on Oracle partitioning for maintaining data for a desired retention period providing performance benefits and purging older data from the database The data retention period for business and market data is configurable Range partitioning of the tables is by date The Data Retention Manager enables you to manage Oracle database partitions and indexes on a daily weekly and or monthly basis Refer to Figure 53 on page 197 This utility allows special processing for trade related database tables to maintain open order execution and trade data prior to dropping old partitions As administrato
456. tenebre eter ent et E E E A A ette deep Do e ep ri 216 Confiunng the Alert Darge UnU see ARTT nero PEU Wee Toate eats 217 Eixecuting the AA ler Purge URB aiite t eee trt pn ttr ip POR GR DOLOR GOES IRAE 219 Procesin JOP FUTON s irte ettet RA Te E tutt nter eL Le teers ea E rre 219 Sample Alert Purge Processes vscsissatsissessssisaseschssstcessessssesvtvadanvsvansesasuesen soseasencestosseucestonvdvansnssavsnceseseavencestvesseneusevezs 221 Sample Purge Alerts Utility Run Situation Oe eese tette tnnt tetntenenn 222 Sample Purge Alerts Utility Run Situation Tuo es eese tette tinent tentent tentnteentetntenent 222 Sample Purge Alerts Utility Run Situation Three eee eee tette tenete tentent tenete 222 Batch Control Diii erEEEE 223 Batches ir Behavior Detectiofhi coo ier erp e teen enne eripi p arie etn ig 223 Directory Structure M R DDEDER 224 MOOS es dsb X 224 Using the Batch Control Utility isin ccsicctscssacssecssccsansvetiovaswsvaeaslecstcdessstasspsetuseats losdisosdevsdcendvebcessbvbensdoansasedondiosatenede 225 Configuring the Batch Control UUU piopi renani E aE EEEE E RE eg bs PR DIeR ede 225 AU IUE 225 Starting a Batch Process Manual eoe a s t die apa dabwdm eme eb e 227 Processing Jor Batob SIgtas udi reve titre eerte i rtt vetet rtt ede ti eei pt detis eee detenta 227 Ending a Bath Processo s
457. ter 8 Administrative Utilities Table 111 Environment 2 Test UAT TABLE_NM SEQUENCE_NAME CURRENT_VALUE MIN_VALUE MAX_VALUE KDD ATTR ATTR ID SEQUENCE 20000000 20000000 29999999 KDD AUGMENTATION AGMNT INSTN ID SEQ 20000000 20000000 29999999 KDD DATASET DATASET ID SEQUENCE 20000000 20000000 29999999 KDD JOB JOB ID SEQ 20000000 20000000 29999999 KDD LINK LINK ID SEQ 20000000 20000000 29999999 KDD LINK ANLYS NTWRK DEF NTWRK DEFN ID SEQ 20000000 20000000 29999999 N KDD LINK ANLYS TYPE CD TYPE ID SEQ 20000000 20000000 29999999 KDD LINK SUMMARY LINK SUMMARY ID SEQ 20000000 20000000 29999999 KDD LINK TYPE SUMMARY LINK TYPE SUMMARY ID 20000000 20000000 29999999 SEQ KDD NODE NODE ID SEQ 20000000 20000000 29999999 KDD NTWRK NTWRK ID SEQ 20000000 20000000 29999999 KDD PARAM SET PARAM SET ID SEQ 20000000 20000000 29999999 KDD PTTRN PTTRN ID SEQ 20000000 20000000 29999999 KDD RULE RULE ID SEQ 20000000 20000000 29999999 KDD SCNRO SCNRO ID SEQ 20000000 20000000 29999999 KDD SCORE SCORE ID SEQ 20000000 20000000 29999999 KDD SCORE HIST SCORE HIST SEQ ID SEQ 20000000 20000000 29999999 KDD TSHLD TSHLD ID SEQ 20000000 20000000 29999999 KDD TSHLD HIST HIST SEQ ID SEQ 20000000
458. th to the target directory where the utility writes extracted metadata INSTALL DIR database db tools data extract backup dir Full path to the target directory where the utility writes backups of the extracted metadata INSTALL DIR database db tools data temp extract prod uct range only Indicator Y or N of whether to extract custom patterns jobs thresholds threshold sets and scoring rules when extracting a scenario Set to Y to prevent extraction of these entities extract prod uct range check For internal use only Indicator Y or N of whether to fail the extraction of a sce nario if any metadata has sequence IDs outside the prod uct range Set to Y to fail the extraction Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 279 Scenario Migration Utility Chapter 8 Administrative Utilities Configuring Scenario Load Table 109 describes scenario load parameters Table 109 Scenario Load Parameters Parameter Description load database user User to use to connect to the database when extracting name scenarios DB_UTIL_USER load database pass Password for the above user word load conn driver Database connection driver that the utility is to use ora cle jdbc driver OracleDriver load conn url Database connection string that the Scenario Migration Utility is to use load ignore cus When set to N custom p
459. the Alert Creation Job To Run Multi match Alert Creator To Run Single Match Alert Creator The Alert Creator is part of the Behavior Detection subsystem Behavior Detection provides default job templates and job template groups for running Alert Creator These jobs can be modified using Administration Tools Refer to the Oracle Financial Services Behavior Detection Platform Administration Tools User Guide for more information The following sections describe running each type of Alert Creator To run the multi match Alert Creator follow the steps 1 Verify that the dispatcher is running 2 Run the start_mantas sh script as follows start_mantas sh 502 where 502 is the job template that Behavior Detection provides to run the Alert Creator algorithm To run the single match Alert Creator follow the steps 1 Verify that the dispatcher is running 2 Run the start_mantas sh script as follows start_mantas sh 503 where 503 is the job template that Behavior Detection provides to run the Alert Creator algorithm Understanding Advanced Alert Creator Configuration Advanced Rules The Alert Creator algorithm can support grouping strategies that the Administration Tools do not support To use these advanced strategies you must enter Alert Creator rules directly into the database The following section discusses these advanced rules The executable retrieves new unowned single matches generated from specified types of scenar
460. the job errors were generated due to incorrect setups you should notify the System Administrator who can correct the problem setups Note The dispatch log may contain a JVM core dump This does not indicate the actual cause of an error you must Refer to the job log for the underlying error To monitor a specific job or to look at the job log history for diagnostic purposes follow the steps 1 Type tail log at the system prompt and press Enter where 10g is the name of the job log file The job log scrolls down the screen 2 Press Ctrl C to stop the display 3 Type lpr job lt job_id gt lt date gt lt time gt at the system prompt and press Enter to print the job log CAUTION This job log file may be a lengthy printout 22 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Clearing Out the System Logs Chapter 2 Behavior Detection Jobs Clearing Out the System Logs Clearing the Dispatch Log Clearing the Job Logs Periodically you need to clear out the dispatch and job log files Otherwise the files become so large that they are difficult to use as diagnostic tools and their size can impact the performance of the system Note Oracle Financial Services recommends that the Oracle client establish a policy as to the frequency for clearing the logs and whether to archive them before clearing CAUTION Before you shut down the dispatcher to clear the system logs veri
461. the start mantas sh script as follows Start mantas sh 505 where 505 is the job template that Oracle Financial Services provides to run the Assignment algorithm Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 177 Auto Close Chapter 6 Post Processing Tasks Auto Close Oracle Financial Services provides a mechanism to close alerts automatically that do not warrant investigation The system can close alerts based on their age status score focus type generating scenario or any combination of these attributes The system regularly evaluates all candidate alerts and closes each alert that satisfies the criteria The system maintains closed alerts for audit purposes and they are still available for display for example from the Entity History page in the Oracle Financial Services UT and processing for example by reopening an alert Defining the Auto Close Alert Algorithm The KDD AUTO CLOSE ALERT table provides all operation sets and their respective operations that the system uses to determine whether it should close an alert The table includes the following Operations are logical expressions that can be used to close alerts for example alert score gt 50 age gt 30 A set of operations based on the same attribute for example score form an operation set e The OPRTN_SET_ID column is a grouping of mutually exclusive operations Each operation specifies the next step th
462. tinued in next page Figure 26 Sample Fuzzy Name Matcher Utility install cfg File Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 121 Fuzzy Name Matcher Utility Chapter 4 Data Ingestion Continued from previous page the mantaslog filename has the format job lt job gt datetimestamp For other subsystems the format is mantaslog datetimestamp Note that category names cannot contain the following reserved words fatal warning notice diagnostic trace category or location db H H Multiple locations can be listed for each property using a comma delimiter log category EXAMPLE CATEGORY location console mantaslog Specify where messages of a specific severity and category should be logged to The valid values are the same as for category Multiple locations can be listed for each property using a comma delimiter If an entry for a severity is not listed here the message will get logged to the location specified for the category by the above property and if that H H db H does not exist it will get logged to the default location configured below log EXAMPLE CATEGORY warning location syslog Specify where a message should get logged for a category for which there is no location property listed above This is also the logging location of the default MANTAS category unless otherwise specified above H Note that if this property is not
463. tion bg If provided runs the tool in the background You can then disconnect your Unix or Linux session without interrupting the tool s operation The system directs any output from the screen to the nohup out file in the directory from which you ran the tool i Uses an input analysis file Table 102 other than the one that input file install cfg specifies Omission of this argument causes the Data Analysis Tool to use the default file in install cfg 0 Writes the output to a file other than the one that install cfg output_file specifies Omission of this argument causes the Data Analysis Tool to use the default file in install cfg Logs The Data Analysis Tool writes status and error messages to the configured log file The default location for this log file is lt install_dir gt database db_tools logs data_analysis_tool log The system writes any system type errors that prevent the tool from connecting to or operating this log file It also writes data errors to the log and includes them in the data analysis report output Refer to Understanding the Data Analysis Report on page 272 for more information Understanding the The tool generates a data analysis report which resides in the location you specified in Data Analysis Report the install cfg file or with the command line o argument 272 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Analysis Tool Chapter
464. tion deploys the Data Analysis Tool as one of the utilities under the database tools Basic configuration for these tools is through the install cfg file that resides in INSTALL DIR database db tools mantas cfg 262 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Analysis Tool Chapter 8 Administrative Utilities To Configure the Analysis XML File Table 100 provides the configuration instructions for the properties that the Data Analysis Tool uses in the install cfg file Table 100 Configuration Instructions for the install ctgFile Property Description Example database driv erName Database connection driver that the utility is to use database driverName oracle jdbc driver OracleDriver urlName schema busi ness owner utils database Database connection string that the Data Analysis Tool is to use Database user for the BUSI NESS schema utils database url Name jdbc oracle oci PROD_DB schema business owner BUSINESS schema market owner Database user for the MAR KET schema schema market owner MARKET dat database username User name for the database The Data Analysis Tool con nects to the database as the INGEST USER for the appro priate privileges dat database user name INGEST USER dat database password Password for the database This is set by the Password Manager Util
465. tion processing can occur on the same day most useful for Best Execution scenarios e Processed data that originates from one source in multiple batches e Delivery of data by an Oracle client can occur in sets throughout the day so that Behavior Detection pro cesses the data as the system receives it This spreads processing over a larger time period Disadvantages e May take significantly longer to process a given amount of data when processing with multiple ingest cycles as opposed to T 1 processing e Applies only to transactional trading and market data e Requires updating of reference tables rather than truncating and reloading them This impacts perfor mance e Becomes difficult to implement and schedule in a mul tiple batch environment Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 77 Alternatives to Standard Data Ingestion Practices Chapter 4 Data Ingestion Table 26 Data Ingestion Options and Alternatives Continued Typical Process Single Source Process that considers all data to reside in one set of source systems for the same time Self containment of data sustains referential integrity across data types Data processing occurs in a single ingestion cycle Advantages e Simplifies configuration e Makes processing easier to monitor than that for mul tiple batch processing Disadvantages e Eliminates support of timely delivery of data
466. to account for any special XML characters contained in the expression for example gt or lt Lookback optiona The number attribute indicates the number of seconds minutes hours days to look back from the current date time to create a time window in which to consider alerts for correlation This is a create timestamp of the alert The wmit attribute identifies the unit of the lookback number Possible values are S M H D and CM for seconds minutes hours days and current month respectively All of these require a valid number value except for CM which essentially just makes the lookback the 1st of the current month for example if the current date is October 14 we will lookback to October 1 if the CM unit is selected The create timestamp of the alert is used to determine whether or not an alert falls within the lookback period Note Do not use a unit less granular than a day in rules intended for batch alerts S M and H are intended for posted alerts For batch processing use D or CM as a unit Scenarios optiona Identifies the Scenario s an alert should have been generated from in order to be considered for a correlation by this rule If not specified system will consider all the scenarios ExistingCorrelationParams required Defines the conditions for extending existing correlations When a new correlation is discovered it is possible that it is a superset with only the triggering alert not already included in the
467. to verify that the format of the incoming Oracle client data is correct and validate its content specifically Error checking of input data Assigning sequence IDs to records e Resolving cross references to reference data Checking fot missing records e Flagging data for insertion or update 60 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Process Flow Chapter 4 Data Ingestion Preprocessors place output files in the directories that Table 20 lists Table 20 Preprocessing Output Directories Directory Name Description inbox lt yyyymmdd gt Backup of input files for restart purposes if necessary data lt business or e Data files for loading into the database as market gt load data type yyyymmdd batch name N XDP e Load control files logs lt yyyymmdd gt Preprocessing and load status and error messages data errors yyyymmdd Records that failed validation The file names are the same as those of the input files data firm transform TC trading data files that the FDT processes Figure 14 summarizes preprocessing input and output directories quei iz Ly Figure 14 Preprocessing Input and Output Directories 3 Simultaneous execution of runDL sh scripts within hardware capacities loads each type of data into the FSDM This script invokes a data loader to load a specified preprocessed data fil
468. tools mantas cfg corrRule cfg Configuring Alert Table 114 describes alert correlation rule extraction parameters Correlation Rule Extraction Table 114 Alert Correlation Rule Extraction Parameters Parameter Description extract database User to use to connect to the database when extracting username alert correlation rules DB UTIL USER extract database Password for the above user password extract conn driver Database connection driver that the utility is to use oracle jdbc driver OracleDriver extract conn url Database connection string that the Alert Correlation Rule Migration Utility is to use extract system id System from which the alert correlation rule was extracted extract schema mantas MANTAS schema owner in the database into which extrac tion of the alert correlation rule occurs MANTAS extract dirname Full path to the target directory where the utility writes extracted metadata INSTALL DIR database db tools data extract backup dir Full path to the target directory where the utility writes backups of the extracted metadata INSTALL DIR database db tools data temp 290 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Alert Correlation Rule Migration Utility Chapter 8 Administrative Utilities Configuring Alert Correlation Rule Load Table 115 describes alert correlation rule load parameters Table 115
469. tory lt INSTALL_DIR gt database golden_data threshold_ sets This directory consists of test threshold sets of all the scenarios that are available with the OFSAAI system 2 Execute shell script 1oad tshld set sh This shell script installs the control data threshold sets for all the scenarios that are installed at your site It also creates new jobs and template group ID s corresponding to all the scenarios installed These template group ID s are same as the scenario ID s of installed scenarios 3 Once the control data thresholds are installed the system is ready for a test run that is generating test alerts Running Behavior Detection on Control Data In otder to generate alerts on the ingested control data execute the new scenario jobs These jobs consists of same template group ID as the scenario ID Refer to Chapter 2 Bebavior Detection Jobs on page 11 to get information regarding about running Behavior Detection Jobs Important Notes 1 Run loaded scenarios with the system date as 20091210 with the following exceptions a For Portfolio Pumping scenario the system date must be 20091204 b For Active Trading scenario the system date must be 20091130 2 Check for system errors in the appropriate logs refer to Appendix A Logging on page 299 for more information Run post processing procedures Close the batch to enable display of alerts in the Behavior Detection UI Log in to the Behavior Detection UI with t
470. tribute value for subClassTagLevel1 should exist in the CASE SUB CLASS LVL1 CD column of the CASE schema table KDD_SUBCLASS1 The attribute value for lt subClassTagLevel2 gt should exist in the CASE SUB CLASS LVL2 CD column of the CASE schema table KDD SUBCLASS2 The CDATA section of the XML tag lt AlertAttrOperations gt is validated as follows 292 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Alert Correlation Rule Migration Utility Chapter 8 Administrative Utilities m The valid operations should be one of BOTH TO and FROM m The valid operators can be gt lt gt IN NOT IN AND and OR m The TO and FROM alert operations can be used only to compare alert attribute values to each other and not to a literal m The FROM alert operation should always precede the TO alert operation m The BOTH alert operator must be used to compare alert attribute values to a literal m The expression can be nested to any arbitrary length provided it confirms to a general syntax Operand Operator Operand Logical Operator Operand Operator Operand For example a BOTH SCORE CT gt 0 b BOTH SCORE_CT gt 0 AND FROM SCORE CT TO SCORE CT Note A space character is expected between each Operand and Operator combination m The precedence of an operation may be depicted using a pair of parenthesis and m The alert attributes provided should be a va
471. trol 1og that the utility generates during execution Logs As the Batch control Utility manages batch processing it generates a date stamped log in the INSTALL DIR database db tools logs batch control 1log file The log file contains relevant information such as status of various batch control processes results and error records You can modity the current logging configuration for this utility in the configuration files lt INSTALL DIR gt database db tools mantas cfg install cfg and categories cfg For more information about logging in these configuration files Refer to Common Resources for Administrative Utilities on page 198 and Appendix A Logging on page 299 for more information 224 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Batch Control Utility Chapter 7 Batch Processing Utilities Using the Batch Control Utility Configuring the Batch Control Utility Setting Up Batches The Batch Control Utility typically runs as part of automated processing that a job scheduling tool such as Maestro or Unicenter AutoSys controls The utility starts and terminates through a shell script using values in parameters that particular configuration files contain The following sections describe this process including tasks that you can perform when configuring the utility or running it manually that is starting stopping or obtaining a batch name e Configuring the Batch Co
472. tructure if it does not already exist e Requires more skilled expertise to monitor and maintain File based Market Data Mechanism for receiving data in flat files from an Oracle client that conform to formatting guidelines in the DIS Advantages e Easier to maintain Lowers total cost ownership TCO An Oracle client does not need a live feed application Disadvantages e Requires longer data preparation times which delays the start of a batch cycle e Increases the costs of hardware and software for timely processing e Makes support of intra day processing difficult due to latency in providing flat file market data 78 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Alternatives to Standard Data Ingestion Practices Chapter 4 Data Ingestion Table 26 Data Ingestion Options and Alternatives Continued Typical Process Truncate Reference Data Process for overwriting non transactional data for example account information customer lists security symbols etc on a daily basis Truncating the data tables before loading them provides a faster mechanism for data loading The Oracle client must provide a complete set of reference data if using this approach Advantages e Supports an Oracle client that does not track updating of daily records well e Provides increased performance when a large per centage of records change on a daily basis Dis
473. ts for all jobs in the chronological sequence that each event occurred To monitor the dispatch log file as it receives entries follow the steps 1 Change directories to the log directory 2 Type tail f dispatch log and press Enter at the system prompt Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 17 Performing Job Tasks Chapter 2 Behavior Detection Jobs The log file scrolls down the screen 3 Press Ctrl C to stop viewing the log file 4 Type lpr dispatch 1log and press Enter at the system prompt to print the dispatch log file Note The dispatch 1og file can be a lengthy printout Performing Job Tasks At the system level the Oracle Financial Services administrator can start restart copy stop monitor and diagnose jobs The sections below cover the following topics e Understanding the Job Status Codes e Starting Jobs e Starting Jobs without the Dispatcher e Restarting a Job e Restarting Jobs without the Dispatcher e Stopping Jobs e Monitoring and Diagnosing Jobs Understanding the Job Status Codes The following status codes are applicable to job processing and the dispatcher The Oracle Financial Services administrator sets these codes through an Oracle Financial Services Job Editor e NEW start Indicates a new job that is ready to be processed e RES restart Indicates that restarting the existing job is necessary e IGN ignore Indicates that the dispat
474. two jobs numbers 300000 and 300001 which relate to scenario numbers 300000 and 300001 respectively As a result of this job the process creates 50 matches and nine alerts and performs nine actions Table 81 defines the matches that relate to these alerts and actions Match ID Range Job ID Scenario ID Alert ID Status Actions Type Date 300000 4 300000 300000 None None 300005 9 300000 300000 300000 OP None 300010 14 300000 300000 300001 OP 300000 OP on 11 6 2006 300015 19 300000 300000 300002 NW 300001 OP on 11 6 2005 300002 on 11 6 2006 NW 300020 22 300000 300000 300003 OP None 300023 24 300001 300001 300003 OP None 300025 27 300000 300000 300004 OP 300003 OP on 11 6 2006 300028 29 300001 300001 300004 OP 300003 OP on 11 6 2006 300030 32 300000 300000 300005 NW 300004 OP on 11 6 2005 and 300005 on 11 6 2006 NW 300033 34 300001 300001 300005 NW 300004 OP on 11 6 2005 and 300005 on 11 6 2006 NW 300035 39 300001 300001 300006 OP None 300040 44 300001 300001 300007 OP 300006 OP on 11 6 2006 300045 49 300001 300001 300008 NW 300007 OP on 11 6 2005 300008 on 11 6 2006 NW Note While the Action ID values are not in time order their impact on the example above is negligible The key aspects of the actions relevant to the discussion are the dates of the actions As a result a range of matches is associated either wholly or partly with an alert an
475. u to query the Behavior Detection Watch List tables to determine if a given name or a name closely matching the given name is on a watch list Refer to the Oracle Financial Services Services Guide for more details on how the service can be called and the results that are returned Watch List Service uses three scripts to start stop and re read the Watch List tables in case of changes These scripts are placed in ingestion_manager scripts folder The scripts are startWatchList sh script to start the web service shutdownWatchList sh script To stop the web service initWatchList sh script this script causes the Watch List Service to re read the Watch List tables in case there have been any changes to it while the service has been running There is a polling agent that runs as part of the web service that will query the Watch List tables on a configurable interval default of 10 seconds This is to done to keep the service in syne with the Watch List Management Utility Alert Processing Web Services The Alert Processing Web service provides the ability to execute additional processing steps during a call to the existing PostAlert service operation currently delivered with Investigation Management Details on this service can be found in the Oracle Financial Services Services Guide Instructions on Administering the Alert Processing Services Alert Processing Service provides two scripts one to start the service and one t
476. ugh Function Role Map function which is available in Security Management System by logging in as the System Administrator in OFSAAI toolkit All Alert Management user roles should be mapped to the function AMACCESS in order to access an alert Users of roles that are not mapped to this function cannot access the details of the Alerts All Case Management user roles should be mapped to the function CMACCESS in order to access a Case Users of roles that are not mapped to this function cannot access the details of the Case This function should be mapped to Case Analyst1 Case Analyst2 and Case Supervisor Roles to assign ownership of a case without applying restriction on the Organization associated with the Case If the ownership assignment is required to be restricted based on Organization associated with the Case for any of these user roles then the RSGNTALL function need not be mapped to the above roles Note The Function to Role mappings ate pre packaged with the solution Defining the User Access Properties and Relationships The following types of data compose a user s security configuration e Business Domain s Property that enables an Oracle client to model client data along operational business lines and practices e Jurisdiction s Property that enables an Oracle client to model client data across such attributes as geographic location or type or category of a business entity 30 Oracle Financial Services Behavio
477. ultiple computer systems Each configuration processes a distinct data stream Configuration spreads Data Ingestion instances across multiple computers Advantages e Enables an Oracle client to scale for higher process ing volume by using multiple computers This increases bandwidth with the use of multiple I O channels e Provides a lower cost structure with several smaller computers than with a single large capacity com puter Disadvantages e Makes configuration and monitoring of the ingestion process more difficult e Requires an Oracle client to split data prior to delivery for Data Ingestion Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 79 Data Ingestion Directory Structure Chapter 4 Data Ingestion Data Ingestion Directory Structure The Data Ingestion subsystem components and data are organized in subdirectories below the ingestion_manager root level Figure 24 illustrates the subdirectories that the ingestion_manager directory contains Additionally Table 27 provides details about each subdirectory ingestion_manager extract transform load extract transform load Scripts SrcFiles repository Contains dated subdirectories Figure 24 Data Ingestion Subsystem Directory Structure Installation of the Data Ingestion subsystem is normally on a single server When requiring high availability or improved performance however installation on two or more serv
478. umn Type Nullable Y N Default PATH ID NUMBER 10 No PATH NM VARCHAR2 50 No QUERY DEF NM VARCHAR2 50 Yes ALERT FOCUS ID KDD CENTRICITY CNTRY ID NUMBER 10 Yes MTCHD TABLE NM KDD EJB NAME EJB NM VARCHAR2 50 Yes BUS NTITY ID KDD CENTRICITY CNTRY ID NUMBER 10 Yes The purpose of this table is to define paths that can be used by the Alert Correlation algorithm to perform the first step in its process correlating alerts to business entities The way such a path is established is by first defining whether the origin of the path Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 187 Alert Correlation Chapter 6 Post Processing Tasks should be an alert s focus or one of its matched records This is established by either populating the ALERT_FOCUS_ID column indicating that the origin should be the focus of the alert or the MTCHD_TABLE_NM column indicating that the origin should be a matched record of the alert The destination of the path the business entity we are trying to correlate to by executing this path is defined by the BUS_NTITY_ID column The actual SQL we would need to execute to establish the relationship between the alert s focus or matched record and this business entity is defined by an element in a query definition file The QUERY_DEF_NM column corresponds to the element name in the query de
479. used to rebuild Institutional Account Summary Month Table from dump file 0160 w_p0160 truncate csms This workflow truncates Customer Summary Month Stage 0170 w p0170 truncate iasms This workflow truncates institutional account summary month stage 0210 w p0210 aggregate bot to file This workflow aggregates sums and counts various fields from BOT for the account summary table It puts the results in a flat file for later processing 0220 w p0220 aggregate fotps to file This workflow aggregates sums amp counts various fields from the FOTPS for the institution account sum mary and customer summary tables It puts the results in a flat file for later processing 0230 w p0230 aggregate deals to file This workflow is used for Daily Aggregation of Deals This populates the Institutional account deal activity file and customer deal activity file 0240 w p0240 aggregate instructions to file This workflow performs Daily Aggregation of Instruc tions This populates the Institutional account deal activity file and customer deal activity file 0250 w p0250 aggregate trades with ca t o file This workflow performs Daily Aggregation of Trades with Corporate Actions This populates the Institutional account deal activity file and customer deal activity file 320 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Informat
480. ut Specifies the communication timeout value in 60 seconds for the MDS when retrieving market summary information from a live market feed MDS MarketHours Specifies the time zone that the live market GMT marketTimeZone feed uses when reporting timestamps MDS MarketHours Specifies the time zone that the local system EST localTimeZone uses MDS DailySummary Specifies the number of milliseconds to wait 100 SubscriptionWait between subscription requests to the live market feed MDS DailySummary Specifies the number of seconds to wait fora 60 LastSubscriptionWait response for all subscription requests before rejecting subscriptions that have not received a response MDS QuoteSizeMultiplier Specifies the number to multiply quote sizes 100 value coming from the live market feed that is the lot size MDS MarketTimeDelay value Specifies the delay in seconds to apply to 30 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 93 Data Ingestion Directory Structure Chapter 4 Data Ingestion Table 32 Datalngest xml File Configuration Parameters Continued DerivedTrade sequence IDs for DerivedTrade records Property Name Description Example MDS HaltedCodes value Specifies status codes within the marketdata ND NP IMB EQP HRS that indicate a halt in trading Ivc T
481. ut down with errors When you are ready to restart the dispatcher and you want to see which jobs had real errors and which jobs generated errors only because they were shut down during processing review the error messages in the job logs For those jobs that shut down and generate errors because the dispatcher shut down a message similar to the following appears Received message from dispatcher to abort job If the job generates a real error a message in the job log file indicates the nature of the problem To view active jobs and then shut down the dispatcher follow the steps 1 Type ps efw grep mantas and press Enter at the system prompt Allinstances of the mantas process that are running appear on the screen Only one instance of mantas should run for each active job 2 Type stop_chkdisp sh lt dispatcher name gt and press Enter at the system prompt This script shuts down the dispatcher Monitoring the Dispatcher The install cfg file that was set up during server installation contains the kdd dispatch joblogdir property that points to a log file directory The log directory is a repository that holds a time stamped record of dispatcher and job processing events Each time the dispatcher starts or completes a job it writes a status message to a file called dispatch log in the log directory This log also records any failed jobs and internal dispatcher errors The dispatch 1og file holds a time stamped history of even
482. ut the Behavior Detection application there are several system accounts that may require changing the password for security purposes Table 18 summarizes the different system account passwords used by the application the subsystems that use those passwords and instructions on how to change the passwords Table 18 System Account Passwords System Account Subsystem Instructions Data Ingest User Data Ingestion 1 Change the password in the database server for this user INGEST_USER 2 Use the Password Manager Utility to change the password This password also needs to be changed on the Informatica pmserver 1 Start the Informatica Workflow Manager 2 Connect to the appropriate repository as the Administrator user 3 Select the Relational option from the Connections drop down menu 4 Click the ingest_user connection that appears under Oracle 5 Click Edit 6 Type the desired password in the Password field 7 Click OK 8 Click Close 9 Select the Save Repository item from the Repository drop down menu Algorithm User Behavior Detection 1 Change the password in the database server for this user KDD_ALG Services 2 Use the Password Manager Utility to change the password Data Miner User Alert amp Case Management 1 Change the password in the database server for this user KDD MNR Data Ingestion 2 Use the Password Manager Utility to change the password Web Application Alert amp Case Manag
483. utility and maintain database partitions and indexes e Configuring the Data Retention Manager e Executing the Data Retention Manager e Creating Partitions e Maintaining Partitions e Maintaining Indexes Configuring the Data The INSTALL DIR database db tools mantas cfg install cfg file Retention Manager contains common configuration information that Data Retention Manager and other utilities require for processing Refer to Figure 54 on page 207 for a sample install cfg file Note The configuration parameters in the install cfg are only used if command line parameters are not provided It is strongly recommended that you provide command line parameters instead of using the install cfg parameters The Data Retention Manager automatically performs system checks for any activity that may result in an error for example insufficient space in the tablespace If it discovers any such activity it logs a Warning message that identifies the potential problem If Data Retention Manager fails to run successfully you can configure the utility so that the ingestion process for the following day still proceeds The following sample section from the install cfg file provides other configuration information specific to this utility including required and optional parameters HHHHHHHHH DATA RETENTION MANAGER CONFIGURATION 1HHHHHHHEHHHHBHHBH E Set the Data Retention Manager input variables here Li drm operation P drm partition type
484. val purposes Clients can now export groups of alerts and cases to a configured location The exported data includes XML metadata for easier searching of the archived documents and attachments associated with the investigation record Exporting alerts and cases ate accomplished in a batch process by running a script Filters can be configured to define the alerts and cases that are exported The batch export process is initiated by running the script run pdf archival utility sh 230 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Batch Export Utility Chapter 7 Batch Processing Utilities Note Prior to using the export functionality you must have at least one alert or case your export directory location must be configured and the Alert Management UI must be running Overview 1 The run pdf archival utility sh script initiates a call to the Investigation Export Web Service 2 Using standard logging procedures the results of the Batch Export are noted to a log file pa archival 1og 3 If expott filter criteria has not been defined for specific alert or case IDs then the system updates the last run date with a date and timestamp at the completion of the process 4 The Web service stores the exported data in the configured directory Expotted data includes multiple file types XML metadata PDF alert case infor mation and other files such as document or spreadsheet atta
485. vel specifies the level of logging detail A level of 7 is the default The command runTIBW sh 1 runs TibWatch with a log level of 1 runMDS sh Launches an instance of MDS runMDS sh The preprocessor terminates after it finishes preprocessing the data that is currently in its memory 82 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Data Ingestion Directory Structure Chapter 4 Data Ingestion Table 28 Run or Stop Scripts by Component Continued Script Names runDP sh lt data type gt Description or Special Instructions Launches an instance of the data preprocessor runDP sh After receiving a soft kill the preprocessor terminates after it finishes preprocessing the data that is currently in its memory If you configure the Preprocessor to run in batch mode you cannot use the stopDP sh script For example runDP sh Customer To run or stop a specific Data Preprocessor specify a valid input component that the run or stop script recognizes If the script does not recognize the input component it exits with an error and identifies the valid list of parameters For valid component names refer to Figure 12 on page 55 Note A Data Preprocessor that you configure to run without polling that is batch mode stops automatically when no data remains for processing However running a stopDP sh script does not terminate batch processing runMDT sh Launches th
486. vices Behavior Detection Platform 6 1 3 Administration Guide APPENDIX E Moving Oracle Financial Services System Environment This appendix describes the steps to move Oracle Financial Services Behavior Detection system from one environment to another Moving PROD Environment to UAT After taking backup of Oracle database follow the steps to move Oracle Financial Services Production environment to the UAT I Move database to UAT using the backup Configure Informatica Refer to Chapter 3 Installing the Application Server in Instal lation Guide for more information on Informatica Update the Informatica counter in KDD_COUNTER table Refer to Installation Guide for more information on updating the KDD_COUNTER table Initialize all Case Sequence present in the Case Schema Setup the Oracle Financial Services Behavior Detection Platform User Interface UD Refer to Chapter 4 Installing Oracle Financial Services User Interface in Installation Guide Set up Oracle Financial Services Alert Management and Case Management Refer to Oracle Financial Services Analytical Applications Infrastructure Installation and Configuration Release 7 3 and Oracle Financial Services Enterprise Alert and Case Management Stage 3 Installation Guide for more information Refer to Scenario Migration Best Practices on page 283 for more information Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide
487. w ph3230 update inst activity risk ORION Common 3191 ommon Informatica Miscellaneous Post Watch List Workflows Fraud Detection External Predecessor Sequence 13 Sequence 14 Sequence 20 Sequence 21 Post Watch List Miscellaneous workflows perform processing in support of workflows in multiple functional areas Figure 47 illustrates Post Watch List Miscellaneous Informatica workflows for Fraud Detection 3210 Watch List Note Processing can occur simultaneously within each sequence Workflows with predecessors must wait for predecessor results before workflow processing can begin Figure 47 Post Watch List Workflows Fraud Detection The application completes these workflows in four sequences processing of a workflow requires data from a Watch List predecessor refer to Informatica Watch List Workflows Fraud Detection on page 160 for more information Refer to the note in the section Informatica Miscellaneous Post Watch List Wor amp flows AMD Brokerage on page 140 for more information Table 64 describes the Informatica Miscellaneous Post Watch List workflows in Figure 47 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 163 Fraud Detection Informatica Workflows Chapter 5 Informatica Workflows Table 64 Post Watch List Workflows Fraud Detection Sequence Number Folder Workflow Number
488. w maintains LOAN SMRY MNTH table 3000 w ph3000 Adjust WL WLS This workflow takes into account the modifications of the WL based on the new user interface WL utility 3005 w ph3005 apply cust KYC risk This workflow finds CUST records that have effective risks that are lower than the KYC risk for the same cus tomer and updates the CUST record s effective risk with the high KYC risk 3010 w ph3010 truncate ls The workflow truncates Link Stage 3020 w ph3020 truncate nms The workflow truncates Name Match Stage 3030 w ph3030 truncate wls The workflow truncates Watch List Stage 3040 w ph3040 truncate wls2 The workflow truncates Watch List Stage2 3051 w ph3051 create addresses from fot This workflow maintains the addresses in the Derived ps Address table It derives the addresses from the FOTPS table 3061 w ph3061 create addresses from INS This workflow maintains the addresses in the Derived Address table It derives the addresses from the INSURANCE table 3070 w ph3070 load watch list staging t This workflow determines changes in the Watch List able table Each entry is classified as Add No Change or Retire based on the comparison of the current day watch list data to the previous day watch list data The output is written to a temporary table 3080 w ph3080 create external entities This workflow maintains the External Entity table It from INS derives the entities from the INSURANCE table 3090 w ph3090 create external entities
489. when one or both of the following occurs e The dispatcher generates errors and stops during mantas processing When the dispatcher is running the Oracle Financial Services administrator can restart a job or jobs by changing each job s status code from ERR to RES A job generates errors and stops during mantas processing If a job stops processing due to errors correct the problems that caused the errors in the job run and restart the job If the dispatcher stops all jobs stop You must restart the dispatcher and restart all jobs including the job that generated real errors To restart a job follow these steps Note If the dispatcher has stopped restart it 1 Type restart_mantas sh lt template group id gt at the system prompt 2 Press Enter When the dispatcher picks up a job from the job queue that has a code of RES it automatically restarts the job Refer to section Starting Jobs on page 19 for more information Note By default the restart_mantas sh script looks for jobs run on the current day To restart a job that was run on a specific date you must provide the optional date parameter for example restart_mantas sh lt template group id DD MON YYYY 20 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide Performing Job Tasks Chapter 2 Behavior Detection Jobs Restarting Jobs without the Dispatcher Stopping Jobs Restarting a job without the dispatcher is
490. who opt to use the queue adapter to process live market data use this parameter eriO MDS Adapter RvSession Daemon The daemon parameter that processing the TIBCO live market feed requires Only Oracle clients who opt to use the queue adapter to process live market data use this parameter tcp 7602 Data Ingest XML Configuration File Table 32 describes the parameters for the DataIngest xml configuration file Caution Default values for properties in this file are suitable for most deployments Use caution when changing any default values Table 32 Datalngest xml File Configuration Parameters Property Name Description Example ProcessingBatch Specifies batch settings that override settings in the database Overrides are primarily useful during testing ProcessingBatch Name Sets the current batch name Ingestion components process only input files that contain this value in the batch name portion of the file name This property should be blank during normal operation ProcessingBatch Date ProcessingBatch Last Sets the current processing date Ingestion components process only input files that contain this value in the processing date portion of the file name This property should be blank during normal operation The date format is YYYYMMDD Identifies the flag that indicates processing of the last batch of the day to Data Ingestion This property should be blank during nor
491. wo types of batches e End of day Represent processing at the completion of a business day for a set of data Some processes are only appropriate for end of day batches For example daily activity summary derivations and calculating alert ages are activities that occur only in end of day batches Multiple end of day batches per day can run if the Behavior Detection installation supports multiple time zones for example New York and Singapore Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 223 Batch Control Utility Chapter 7 Batch Processing Utilities e Intra day Used when loading data between end of day batches to obtain more frequent detection results For example running a batch of trading compliance scenarios at 10 00 A M can identify behaviors relevant to the opening of the market without waiting for the end of the day to be able to act Directory Structure Table 82 provides the directory structure for the Batch Control Utility in INSTALL DIR database db tools Table 82 Batch Control Utility Directory Structure Directory Contents bin Executable files including the start mantas batch sh end mantas batch sh and get mantas batch sh shell scripts lib Required class files in jar format mantas cfg Configuration files for example install cfg and categories cfg in which you can configure properties and logging attributes logs File batch con
492. workflows in six sequences processing of workflows for some sequences requires data from external predecessors refer to Informatica Miscellaneous Optional Workflows AML Brokerage on page 136 and Informatica Watch List Workflows AML Brokerage on page 138 for more information These workflows may proceed simultaneously in all sequences as soon as processing of any predecessors has completed Table 49 describes the Informatica Summary workflows in Figure 32 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Guide 143 AML Brokerage Workflows Chapter 5 Informatica Workflows Table 49 Summary Workflows AML Brokerage Workflow Sequence Number Folder Number Workflow Name Predecessor 16 ORION Production 0150 w p0150 truncate asms ORION_Common 2005 16 MLM Brokerage P 0160 w p0160 truncate csms ORION Common 2005 roduction 16 MLM Brokerage P 0170 w p0170 truncate iasms ORION Common 2005 roduction 16 MLM Brokerage P 0210 w p0210 aggregate bot to file roduction 16 MLM Brokerage P 0220 w p0220 aggregate fotps to file roduction 16 MLM Brokerage P 0230 w p0230 aggregate deals to file roduction 16 MLM Brokerage P 0240 w p0240 aggregate instructions to file roduction 16 MLM Brokerage P 0250 w p0250 aggregate trades with ca to file roduction 16 MLM Brokerage P 0260 w p0260 a
493. xi setting up 46 System Administrator xxi usets 31 rules advanced alert creator configuration 174 running alert creation 174 S Sample install cfg File 207 sample logging configuration file 304 scenario group usets 31 Scenario Manager 46 account setup 46 database access 45 Job and scenario editors 4 logins 45 scenatio metadata extracting 281 Scenario Migration Utility 261 275 configuring 275 277 extracting metadata 281 loading metadata 282 logs 275 scenario extraction 279 scenatio load 280 using 275 scoring alert 172 sctipts env sh 85 firm analyze sh 69 matket analyze sh 69 process firm summary sh 69 process market summary sh 69 run stop scripts 82 runDL sh 67 runDP sh 57 63 runFDT sh 64 runMDS sh 63 runMDT sh 64 runRebuildIndexes sh 69 set mantas date sh 57 statt Wkflw sh 133 Start Batch 8 Status Codes 18 ERR 18 FIN 18 job 18 RES 18 RUN 18 subdirectories backup 105 107 bin 82 config 85 data 103 errors 103 extract 104 105 firm 105 in 106 inbox 106 jars 82 load 105 106 logs 107 market 104 sctipts 82 tibspool 106 transform 104 106 subsystems Behavior Detection 3 8 data ingestion 3 summaty AML banking 151 AML Brokerage 143 170 Broker Compliance 155 summary workflows Fraud Detection 164 Suppress Alert Algorithm 183 suppression alert 172 Oracle Financial Services Behavior Detection Platform 6 1 3 Administration Gui
494. y file that provides directory and network access rights to the component NUM SPLIT FILES Specifies the degree of parallel processing for Informatica ingestion The default is 10 the maximum is 10 NUM SPLIT LINES Uses this parameter during the fuzzy name matching process Behavior Detection splits names into multiple files Although this is parameterized Behavior Detection does not make this parameter transparent to the client The best number of records is determined to be 50000 PROCESS BANK TO BANK PROCESS FOREIGN FL Enables ingestion to derive the BANK TO BANK field Set the value to N if the client provides this field Enables ingestion to derive the PROCESS FOREIGN FL field Set the value to N if the client provides this field server Instructs Java JRE to optimize for server based processing Xms lt NNNN gt Indicates the minimum number of megabytes as NNNN to reserve for Java memory allocation Xmx lt NNNN gt Indicates the maximum number of megabytes as NNNN to reserve for Java memory allocation Note Setting Xmx to too small a size may result in component failure ACCT TRUST FROM CUST Indicates whether the account risk should be exempt or trusted based on the exempt or trusted status of the customer s risk The default value Y Default values that are appropriate to the operating system in use for example Linux or Solaris are automatically set in the env s
495. ystem processes in a set of files after processing the set of files that contains the corresponding New or Cancel or Replace event Such an event that has a timestamp prior to the timestamp of the last event against that order or trade is considered an out of sequence event For example File Set 1 contains the following events e NW order event timestamp 09 30 00 e MF order event timestamp 09 45 00 File Set 2 contains the event MF order event timestamp 09 40 00 This second MF event is considered out of sequence This also applies to trade execution events against orders For example File Set 1 contains the following events e NW order event timestamp 09 30 00 e MF order event timestamp 09 45 00 File Set 2 contains NW trade execution event references the above ordet timestamp 09 40 00 This trade execution event is considered out of sequence It is important to note that this also includes market data If in a given batch market data up to 10 00 00 is used to derive attributes for a given order any event in a subsequent file against that order with a timestamp prior to 10 00 00 is considered out of sequence An out of sequence event has no effect on the order or trade that it references Processing sets the out of sequence flag for the event to Y es and the system writes the event to the database Out of sequence indicators for any summaries that the event affects ate set to Y es which indicates that potential compromise o

Download Pdf Manuals

Related Search

Related Contents

秋穂幼稚園・大海保育園・秋穂保育園 3園の子どもたちが交流しました。  GCU 102 - Arizona State University  Massive Pedestal/post 16317/47/10  EDS-405 User`s Manual  Axis M1025  Samsung SGH-Z720 Manuel de l'utilisateur  Eastern Bank TreasuryConnect Electronic Report Delivery (ERD  point re-por-0810-02.indd  PLANCHA A VAPOR /  Smart monitor 誕生  

Copyright © All rights reserved. Failed to retrieve file