AN ARCHAEOLOGICAL NETWORK DATABASE MANAGEMENT
Contents
1. as this word is used in database terminolgy to avoid confusion it will not be used here However the definition of an object is not confined to artefacts but also features sites and J regions It is anything which is unique and can be identified by a unique code and can be studied in isolation Therefore a sherd of pot can be considered as a single object as can a complete jar while a particular site such as Mucking in Essex is an object in that it can be studied separately from the site of Charlton in Hampshire An archaeological attribute is some facet of an object which is to be described An object usually has many attributes OBJECT gt gt ATTRIBUTE where gt gt has a 1 to many relationship with It is the attributes which are used to distinguish one object from another For example one attribute of a complete jar is that it is complete whereas the sherd is part of an incomplete vessel However if the sherd is part of a rim then both objects can have rim diame ters as an attribute which makes them directly comparable With the two sites of Charlton and Mucking one attribute which refers to both is period another is area and others will involve types and quanti ties of artefacts and features Archaeological data are actual values of attributes and refer to a specific object For an object any single attribute will normally have only one value OBJECT gt ATTRIBUTE gt VALUE wh2. the database and is decided by the programmer and the users of the database 5 An Example The Mucking excavation data Catton et al 1982 has been used to test this database structure There are several large sets of data which have been recorded on to floppy disks and these are being integrated by loading them into a single database file This will enable questioning of the data about the site to be carried out much more easily than having to do it manually Several extension programs have been written to take into account different sets of data stored in the database as well as allowing the use of different output dev ices such as a digital plotter 6 Conclusions This work began in 1981 when 8 bit microcomputers were the most widely available with considerable amounts of software available for them However more recent developments in computer technology are making these systems look out of date especially as the new 16 bit microcomputers are now being sold in the same quantities Thus in the future problems with the amount of memory available for a process will no longer exist especially when the memory is of the order of mega bytes rather than kilobytes Since a version of M D B S exists for 16 bit microcomputers it is hoped to upgrade for these microcomputers the system described here 98 References Catton J P J Jones M U and Moffett J C 1982 The 1965 1978 Mucking Excavation Computer Database Grah
3. AN ARCHAEOLOGICAL NETWORK DATABASE MANAGEMENT SYSTEM Jonathan C Moffett Institute of Archaeology London ABSTRACT This paper describes a Network Database Management System N D B M S It has been designed to be used with data stored using the Simple Data Entry and Retrieval S D E R suite of programs which are described else where Catton et al 1981 Moffett 1984 However it is also possible to use it on its own The N D B M S is based on a general model of archaeologcial data which has a direct equivalent ina network database structure This paper describes this model and describes the D B M S operation 1 Introduction Archaeology is a discipline concerned with large quantities of data obtained by a variety of data recovery techniques such as exca vation survey and attribute analysis All the data that are obtained are in one of three forms numbers e g dimensions and percentages characters e g descriptions and documents images e g photographs Most archaeological research is involved in the study of these three types of data to aid our understanding of what has happened in the past Computers can handle large quantities of data and are capable of handling the three types of data referred to above Thus archaeology Should be a natural beneficiary of computer technology for both storage and analysis 2 Archaeological Data Archaeology is primarily concerned with the study of objects or entities but
4. am I and Webb R E Eds Computer Applications in Archaeology 1981 pp 36 43 e M D B S 19981 User s Manual and Documentation Moffett J C 1984 The generation of archaeological excavation databases in Insti tute of Archaeology Bulletin 98 Figure 1 A General Data Mode attributes ATTRIBTE Key to Relatlonsh ps 1 to fee Many to Many Figure 2 A General Network Database Structure descr Ips f 1dnums JNUMERCAL Key to Names FIELD Record names fields Set names Idents 100
5. at the record DESCRIPT is pointed to itself by the set dindex This is to allow the use of synonyms so that the owner set is the most commonly used description but others which are also used are also stored in the database thus forming a data dictionary 4 2 Editing This option allows the user to modify the contents of the qata base in some way This includes the adding deleting or changing of fields descriptions numerical values or objects 4 3 Listing This allows the user to list information about the database such as listing all the descriptions with or without synonyms and the types of object contained in the database pF 4 4 Statistics This option allows the user to find out how much data is actually stored in the database in order to monitor its size 4 5 General options There are four general options which are available with most menus of the d b m s The first is to set the output destination which can be either the console screen the printer or a file on the disk At most levels there are help commands which display the contents of a help file on the screen It is possible to move to other modes as well as leave the d b m 8 4 6 Extensions The above describes the general options available It is also possible to add other programs to the d b m s hierarchy to do specific tasks and control specific output devices such as a digital Plotter or graphics terminal This really depends on the contents of
6. ent paper represents something of an update of that paper M D B S has a data definition language d d 1 which allows the user to specify the database structure Data are stored in fields called items by M D B S and these are grouped together into records It is possible to have several different record types that is records with different groups of items in the same database file The records are related to each other by sets which are assumed to have an owner and a member and ownership can be in one of the fol lowing four forms and are shown with their diagramatic equivalent ee gt One to one One to many Many to one lt gt Many to many lt lt gt gt gt gt It is also possible to define the order of the member records with respect to its owner record such as numerically or alphabetically sorted last in first out LIFO or first in first out FIFO With the many to one and many to many sets it is possible to define the order of the owner records with respect to a member record There is a data manipulation language d m 1 which can be embedded in a number of computer languages These allow the user to get at the data in the database and manipulate it as required and which takes care of finding the data in the database file Thus the programmer only needs to know what data is wanted and then issue the relevant instructions for the d m l to do the rest The database file is organised in
7. ere gt has a 1 to 1 relationship with Since an object has several attributes it therefore has several values associated with it Rearranging the diagram above it is possible to say that an object has several values derived from it and that each value refers to one attribute OBJECT gt gt VALUE lt ATTRIBUTE Since each value is either a number a character or group of charac ters or an image then it is possible to produce the following struc ture OBJECT NUMBER CHARACTER IMAGES ATTRIBUTE It is possible to generate a network model using these three definitions Given a single object there are many values referring to it of either of the three types of data but each value refers to only one attribute although there are many attributes However given 94 several objects one value might refer to more than one object for instance the same rim diameter may refer to several objects while one attribute may have more than one possible value Figure 1 illustrates this more complex arrangement 3 A Network Database Structure Having developed a general model for archaeological data it is possible to develop a database structure to mirror this but first it is necessary to briefly describe the d b m s used The database management system used in this research is M D B S 1 which is a commercial network d b m s M D B S 1981 One use of this package has been described in Moffett 1984 although this curr
8. o records together and those which are owned by the system These sets are entrances into the database thus in figure 2 the sets fields and idents are both entrance sets and have to be traversed before the sets fldnums or nums can be used 4 A Database Management System The database management system is written in Pascal with M D B S 1 data manipulation commands embedded within it The d b m s is actually made up of about 30 programs which can chain together in predefined sequences depending on the command entered by the user The user is not aware of this division except possibly for the short Gelay as one program replaces another in the memory of the computer The reason for this division is mainly due to considerations of computer memory There is only 64 Kilobytes of memory available to the computer into which not only the program must fit but also the pages of the database Thus the longer the program the less memory there is available for the database and so the greater the number of disk read and writes necessary to search through the database to find a given item Therefore to maximise the amount of memory available for the data and thereby quickening the search time the programs have been made small One other advantage of this is that it allows greater con trol over the individual programs and speeds up compilation time Should any changes be necessary in any one program The programs are mostly menu dri
9. pages of a size specified by the programmer When a program uses the database it reads pages of the database into the memory and writes them back when it has finished However the number of pages that can be put into the memory at any one time depends on the amount of memory available With large data base files the program has to keep reading blocks of pages into memory until it locates the information it requires The file can be spread over several disk drives the d m takes care of any drive changes which may be necessary Figure 2 shows the M D B S database structure used to model the 03 archaeological model described above It is not possible to store images using the computer system which is available so this type of data 1s not considered in this d b m s The database structure modifies the original model to aid the d b m s There are five record types in the database 1 RECORD IDENTNO object identification ITEM object id code 2 RECORD NUMERCAL numerical ITEM value 3 RECORD DESCRIPT description ITEM descriptive value 4 RECORD COMMENT comment ITEM character array 5 RECORD FIELD attribute type ITEM attribute description There are two types of textual data descriptive comments which are usually single word descriptions or abbreviations and longer more ver bose sentences The field record is here used to refer to the attri bute type There are two types of set used by M D B S those which link tw
10. ven Whenever the user is presented with a choice of options a list of these is displayed The d b m s can only be entered by one program called the entrance pro gram which asks for the user s name and password These can be 96 predefined before the database is loaded or added to the database at a later stage It is not possible to enter the database by any other Program in the hierarchy If this is attempted the program in question automatically chains to the entrance program Once the uger has successfully entered the d b m s they are presented with the following menu Database Management Control Menu Choose option Search database Edit database List information about database Database contents statistics vaur l l Set output destination e lt HELP 13 Leave database A D 1 3 4 1 Searching This option allows the user to find objects which fulfill certain conditions It is possible to specify either descriptions and or numerical values in the conditions Two logical operations are allowed AND and OR for intersection and union of sets With numerical values it is also possible to specify a combination of less than greater then and equal to in the condition statement Once the statement has been entered it is checked to see whether it is correct If a description is not found the user is able to choose from the seven closest descriptions in spelling It will be noticed in figure 2 th
Download Pdf Manuals
Related Search