Home

User`s Manual - Knowledge-Based and Software Systems

1. Project title Project acronym Project number Enabling Communities of Interest to Promote Heritage of European Regions Cipher IST 2001 32559 Partners The Open University UK Project coordinator Dublin Institute of Technology Ireland University of Art and Design Helsinki Finland The Discovery Programme Ireland Czech Technical University Czech Republic Internet Loesungen und Dienstleistungen RiS GmbH Austria Report title DNAT User s Manual Authors Jan Uhl Petr Kfemen and Lubo Kral E mail uhlirClabe felk cvut cz Deliverable D26 2 Version 1 0 Report availability Internal Date September 2004 Contents DNAT Us rs Mantal seses ternir eS oleate ennn reastas ieena Sa a E OaE a a A AS A 3 IsGeneral Descrip omran asees e e e EE e E a E a aai 3 2 Installation Instructions edinen ie iaa e e ia eaa a aeae araara asai ai 3 SPA DESEN DIN VA I oy ce sore toy Fes ead T E AE E T 4 Il Creatin DONOSTIA 4 3 2 Description of available fUNCHIONS oooconnnnococcnonocncononocccononannnononnononnonancnnnno cnn ronnnnnccnnnn nn ncnnannnccnanoss 8 33 Why to Use DNAT Po Ad aa e Ad ta 10 IS MN sss isk oasis ob Soe sek sus pags Sausage be I AE EE EEE saved gestae EARE E E EE Ra 10 DNAT User s Manual 1 General Description Dynamic Narrative Annotation Tool DNAT was designed to support users in creating printable knowledge intensive content and the corresponding kno
2. 3 Ontologies pfAlan Turing was brilliant mathematician who mastered che Ten ded Defaut marking marking 3 El cien cracking of the German Enigma code during the World War II 3s Acts and states i A person Ss E amp cipher crm in Bletchley Park England which helped Britain to prevent 5 German invasion 2 z Z a Repository View Developing his key ideas the concept of the universal Conceptual Marking E Area machine its realization by means of electronic technology Area al C actor only an injury prevented his serious considerations for the C group British team in the 1948 Olympic Games y C person ro TERRA R belongs to Conceptual graph module R has contact point AS R has note Aturing R has ownership of R has residence R has type R is identified by C appelation Ontology View Area Te R has note R has type R is identified by C place C temporal entity time specification R belongs to R has note R has type R is identified by C crm meta entity C primitive value C rock machine Conceptual Graph Area CaiProjectsiDNAT_Sample_Filesinarrativetturing xrrl Figure 1 DNAT GUI Layout While annotating a narrative user seeks for important concepts in a narrative text When some words are selected in the annotated text they can be dragged over a CG area of DNAT and dropped there as shown in Figure 2 DNAT will
3. Project homepage o laboratory A http labe felk cvut cz uhlir DNATWeb DNAThome html Ge rstner y C Czech Technical University 10
4. Sets concept add mode A new concept can be placed by single mouse click in the graph Sets relation add mode A new relation can be placed by single mouse click in the graph Sets connection mode A new connection between a concept and a relation can be established by pressing mouse on a concept and releasing on a relation Sets co reference link mode A new co reference link between two concepts can be created with this tool Sets context create mode In this mode new context is created by selecting desired cells Creates new narrative document The same function is available through menu item Narrative gt New Document Opens narrative document either in RTF or in saved XML narrative format The same function is available through menu item Narrative gt Open Document Saves narrative document into XML format The same function is available through menu item Narrative gt Save Document Saves narrative document under different name in XML format The same function is available through menu item Narrative gt Save As Document Selects bold font Selects italic font Selects underline font Aligns current line to the centre Opens ontology in any of these formats XML meta The same function is available through menu item Ontology gt Open Saves current ontology into any of these formats XML meta Common LISP cls CLOS OCML OIL RDFS The same function is available through menu item Ontology gt Save Crea
5. by using Xmx switch of JVM You can also get rid of standard output messages by using javaw exe rather then java exe Example JAVA_HOME java Xmx350m jar Dna jar en En 4 Consult project webpage for getting help 3 Using DNAT 3 1 Creating annotations Basic features and GUI layout of DNAT are outlined in Figure 1 Before a narrative can be annotated user must open at least one ontology in the DNAT The repository view area of the DNAT shows the hierarchy of both opened ontologies and knowledge bases while the ontology view area shows just one ontology that is selected in the repository view area Selected ontology is displayed with all classes inherited from its super ontologies Conceptual Graph Drawing Operations Select Concept Relation Connect Link Context Conceptual Graph Operations New Save Save to Image Open D namic Narrative Authoring File Conceptual Graph Narrative Ontologies Query Builder Time Line Toolbars WERE Gy KB bletchley park kb C crm entity C dimension 9 C existence Narrative Operations New Open Save Save As Ontology KB Operations Open Save Basic Text Editor Functions and its utilization fo lof logical Narrative Area process he became the mm computer principle He was an eccentric character who did not hide his homosexuality After the war he became zealous cross country runner and
6. ontology editing environment created and maintained by the Information and Knowledge Based Systems Group of the Gerstner Laboratory at the Czech Technical University in cooperation with the Knowledge Media Institute at the Open University Milton Keynes Export from internal Apollo CH format to other formalisms is provided by means of export plug ins to Apollo CH which work in DNAT as well Resulting annotation can be exported into many different formalisms intended for semantic annotations supported by DNAT export to OCML RDF DAML OIL and OWL is currently available by means of VO Plug ins When appropriate conceptual type or the entire domain ontology is missing an instance of the most general concept entity is created Such instance can still play role in many conceptual relations and its specific conceptual type can be defined later Therefore the annotation graph remains valid though semantics of annotated content is captured with less accuracy amic Narrative Authoring File Conceptual Graph Narrative Ontologies Query Builder TimeLine Toolbars ace Al Bw v ionospaced 9 Ontologies a KB 1turing kb 5 O Acts and states cipher crm 2 2 Developing E machine its and its utili process he principle He was an ecca KB bletchley park k
7. serious considerations for the British team in the 1948 Oly He was arrested and came to tri on 31 March 1952 after the police learned of his sexual relationship with a young Manchester man He had to und figo one year estrogen therapy in order to prevent being imprisoned KB cipher sbs kb C crm entity C crm meta entity primitive value family relationship C Day Type C Month Name Type C Month Type o Time Granularity C Time Quantity C Time Uncertainty C Week Day Type Alan Turing was found on June 19 in his home with half eaten apple lying beside his bed His death was classified as suicide by cyanide poisoning a Conceptual graph module turing Alan Turing C Documents and Settings uhliriProjects plakatituring xml Figure 2 Identifying Concepts in Narratives In the next annotation step user finds an appropriate conceptual type in the ontology browser of DNAT and assigns it to the newly created concept by means of drag and drop see Figure 3 By doing so the type field of the concept node is set In this moment the drag and drop event triggers automatic creation of an instance of the selected conceptual type in the frame based knowledge base Instances stored in the frame based knowledge base can be seen in the ontology view area of DNAT Slots defined in ontology can be used to
8. 4 Dynamic Narrative Authoring O niki File Conceptual Graph Narrative Ontologies Query Builder Time Line Toolbars ARA bijug k E 4 gt El Del P os pios g Ontologies y Alan Turing was brilliant mathematician who masterminded the cracking of the German Enigma code during the World El a O ciphercrm War ll in Bletchley Park England which helped Britain to prevent German invasion 3 O cipher story and narrative 5 O cipher sbs O Time_CTU Developing his key ideas the concept of the universal machine its realization by means of electronic technology KB cipher sbs time kb and its utilization for computing any kind of logical process he became the father of the modern computer principle 5 KB cipher sbs kb KB turing kb y He was an eccentric character who did not hide his homosexuality z 2 ba KB cipher sbs kh 9 C crm entity C dimension 9 C existence o Cyst oC group AL C person After the war he became zealous cross country runner and only an injury prevented his serious considerations for the British team in the 1948 Olympic Games He was arrested and came to trial on 31 March 1952 after the police learned of his sexual relationship with a young Manchester man He had to undergo one year estrogen therapy in order to prevent being imprisoned Alan Turing was foun
9. b 9 C crm entity e C dimension 9 C existence C actor gt C group homosexuality tric character who did not hide his his key ideas the concept of the universal realization by means of electronic technology zation for computing any kind of logical became the father of the modern computer Default marking person C person Annotation toal R belongs to R has contact point R has note R has ownership of R has residence R has type R is identified by C appelation C contact point C legal object e C stuff gt C type R belongs to B has note B has type e al entity C time specification R helongs to R has note R has type R is identified by C crm meta entity C primitive value C rockemachine place Bletchley Park CaProjectsiDNAT_Sample_Filesinarrativetturing xml Figure 4 Pre selecting the most often used conceptual types Conceptual graph module Pe person s person Alan Turing has residence a all relations visible relations has dimension is subject to has keeper has condition has section definitio has gender is member of legal be has part has reside bears feature right held by has residence place Bletchley Park Figure 5 Connecting Concepts in Annotati
10. d on June 1954 in his home with half eaten apple lying beside his bed His death was classified as suicide by cyanide poisoning a Conceptual graph module turing T Albrecht lil of gk T AlbrechtIv T Alzbeta daughter or T Anna of Rozmital T Antonio Cometa T Antonio Ericero T Antonio Melani T Antonio Vogarelli T Amost ofMannsfeld T B Roubalik T Baltazar Marradas T Baltazare Maggi T Bartolomeus Soller T Bavor lll of Strakonice T Bechyne of Lazany T bishop ofPasoy T bishop Tobias of Bechyne T Bohunka of Pernstejn T_Dahusiaw Mal hal z person Alan Turing Figure 3 Assigning Conceptual Type Conceptual relations related to the selected concept i e represented as slots of ontology concepts can simply be selected from the table of all available slots and then connected to other concepts in the graph This is visualised in Figure 5 When user wants to delete a concept from the annotation graph it is deleted from the corresponding frame based knowledge base as well Thus instances of concepts are created in the knowledge base and visualised in the annotation graph CG node represents a concept used in the narrative and at the same time associates a concept from the narrative with a formal definition of this concept possibly instance in ontology We keep both ontologies and knowledge bases in Apollo CH format Apollo CH is an
11. describe semantic relations i e extrinsic relations between concepts used in a narrative and other concepts that are not explicitly stated in the same narrative DNAT allows yet making these extrinsic relations explicit in a narrative annotation written in CG as they may carry contextual information important for story understanding DNAT provides a conceptual marking feature that simplifies previous two annotation steps The User first selects several conceptual types that are most often used during the annotation process It can be done by dragging a concept from the ontology view area and dropping it in the conceptual marking area which appears on the right relative to the narrative area Different colour can be selected for each conceptual type in the conceptual marking area of DNAT By pre selecting one conceptual type in the conceptual marking area the user defines both a conceptual type and a colour for each term that is highlighted in the narrative area and dropped in the conceptual graph area Thus each term highlighted in the narrative area gets colour of the selected conceptual type and each new instance dropped into the annotation graph has already defined its conceptual type This simplified procedure can be used until the user pre selects different conceptual type or the pre selection is cancelled by selecting Default marking option from the conceptual marking area For better idea about the conceptual marking feature of DNAT see Figure
12. on Graph 3 2 Description of available functions Functions available in both DNAT and its modules are outlined in Table 1 for those functions accessible through the main application menu and in Table 2 for those functions accessible also through application tool bar _Menu items Icon Description File Open Project NOT IMPLEMENTED YET Save Project NOT IMPLEMENTED YET Exit Shuts down the application Conceptual graph New CG Creates new conceptual graph Save CG Saves current conceptual graph The binary format is provided Export to Image Ei Exports conceptual graph to the image Available image formats depend on the used JDK Open CG m Opens saved conceptual graph The binary format is provided Close CG Closes current conceptual graph Rename CG Opens rename dialog for current conceptual graph Narrative New Document E Creates new narrative document Open Document rr Opens narrative document either in RTF or in saved XML narrative format Save Saves narrative document into XML format Save as m Saves narrative document under different name in XML format Ontologies Open Ontology es Opens ontology in any of these formats XML meta Save Ontology pl Saves current ontology into any of these formats XML meta CommonLISP cls CLOS OCML OIL RDFS Close Ontology Closes current ontology Query Builder Configure Server Configures server for the query b
13. tes new time line Opens saved time line Saves current time line Table 2 Toolbar items description Plug in menus and toolbars are written in italic these are shown only in case that the corresponding plug in is loaded 3 3 Why to Use DNAT e Easy to learn easy to use Experiments proved that people are able to read Conceptual Graphs and the knowledge editors who write semantic annotations as Conceptual Graphs can learn to write it in a few hours because with DNAT the process is similar to using colour markers for highlighting key concepts in documents DNAT guides users through the process of connecting concepts by offering context sensitive sets of applicable semantic relations DNAT users need not to be familiar with syntax of semantic mark up languages e Automated complex computer oriented representation tasks During the knowledge extraction process DNAT automatically builds a knowledge base which can be translated to many knowledge representation formalisms such as RDF OWL OCML etc After the knowledge has been translated to CG with human assistance the unambiguous semantic interpretation of the text was provided in a formal and explicit way It was achieved through the assignment of conceptual types with well defined semantics to terms and their interlinking via conceptual relations Therefore the semantic ambiguities are eliminated by the system which enforces a single definition for every term As a result the s
14. then automatically draw concepts as nodes of the annotation graph Each selected word appears in the referent field of a newly created node It means that the instance of the most general conceptual type called Entity was created However some ontologies may use both different name and different structural properties for the most general conceptual type or they may have several concepts at the most general level Therefore in this annotation step the newly created node does not automatically appear in the frame based knowledge base that DNAT maintains besides the annotation graph Dynamic Narrative Authoring 3 E lolx File Conceptual Graph Narrative Ontologies Query Builder TimeLine Toolbars puts elo Ontologies O cipher crm KB turing kb E Cold Rie Bie b ilu v pias As brilliant mathematician who masterminded the cracking of the German Enigma code during the World Ap chley Park England which helped Britain to prevent German invasion q Alan Turing Ontologies Akey ideas the concept of the universal machine its realization by means of electronic technology and its utilization Ry computing any kind of logical process he became the father of the modern computer principle He was an eccentric cfm Time Line acter who did not hide his homosexuality After the war he became ze pus cross country runner and only an injury prevented his
15. uilder Time Line Execute Query Executes timeline query New Time Line 7 Creates new time line Open Time Line gt Opens saved time line Rename Time Line Renames current time line Close Closes current time line Save Saves current time line Save as Saves current time line under different name New Event Creates new event in current time line Edit Event Opens edit dialog for current event Delete Event Deletes current event Sort by Name Sorts events by name Sort by Date Sorts events by date Sort by Type Sorts events by type Table 1 Menu items description 8 Toolbar Conceptual graph Conceptual graph elements Narrative toolbar Ontology toolbar Timeline toolbar Icon a R Pp N T ld le En nu Bu b z U DLO Description Creates new conceptual graph The same function is available through menu item Conceptual graph gt New CG Opens conceptual graph saved in the binary format The same function is available through menu item Conceptual graph gt Open CG Saves current conceptual graph The same function is available through menu item Conceptual graph gt Save CG The same function is available through menu item Conceptual graph gt Export to Image Exports conceptual graph to the image Available image formats depend on the J2RE used Sets selection mode The selection can be made by pressing and dragging mouse in the graph
16. wledge base at the same time DNAT allows creation of semantic annotations in Conceptual Graphs CG CG developed by John Sowa is a human readable notation for First Order Logic FOL DNAT enables users to link text fragments in the narrative and ontology classes to create instances in the CG and the corresponding knowledge base Then CG represents easily readable model of narrative and shows the dependencies of narrative primitives The design structure of DNAT is strictly modular It means that every piece of functionality is provided by a single plug in Therefore for full annotation power following plug ins are needed CG module cg jar Ontology module onto jar Narrative module narrative jar and Marking module marking jar 2 Installation Instructions 1 Download and install Java version 5 0 J2SDK 5 0 RC 2 Download actual release of DNAT and unpack it into any directory 3 Edit rundna bat before running DNAT Batch file rundna bat can be found in your DNAT installation directory a Set JAVA_HOME environment such that it points to your private Java Runtime Environment of your Java 5 0 installation Example set JAVA_HOME c j2sdk1 5 0 jre bin b Change cd command such that it will change the current directory to your DNAT installation directory Example cd c Documents and Settings uhlir cvswork Cipher Dna c The last command shall run DNAT You can increase heap size used by the Java Virtual Machine JVM
17. ystem can automatically translate CG to and from logic and various computational languages e Document annotations People can read CG without special training and a computer system can translate it automatically to different natural languages when multilingual ontologies are available Thus CG can be used for semantic search and reasoning regardless the language of documents used for extracting the knowledge e Multilingual presentation abilities The Conceptual Graph output from the DNAT can also be presented directly as an annotation to the original source documents in a preferred language They can serve as humanly readable comments This is only limited by availability of multilingual ontology which provide basis for the annotation process e Versatility in knowledge representation DNAT allows the annotation to be represented in frame based formalisms e g OCML CLOS semantic mark up languages e g RDF OWL and logic based languages such as CG The open plug in architecture of DNAT enables developers to easily extend its I O capabilities by means of other plug ins Acknowledgement This work is funded by the CIPHER project Enabling Communities of Interest to Promote Heritage of European Regions supported by the European Commission IST RTD program POWERED BY 1ruer DNAT was created in the Information and Knowledge Based Systems Group of the Gerstner Laboratory Department of Cybernetics Czech Technical University Prague

User`s Manual - Knowledge-Based and Software Systems

Contents

Download Pdf Manuals

Related Search

Related Contents