TBMS: Domain Specific Text Management and Lexicon Development

The definition of a Text Base Management System is introduced in terms of software engineering. That gives a basis for discussing practical text administration , including questions on corpus properties and appropriate retrieval criteria. Finally, strategies for the derivation of a word data base from an actual TBMS will be discussed. l. Introduction Textual data are a sort of complex data object that is of growing importance in many applications. Research projects from such different fields as history, law, social sciences, humanities and linguistics but also commercial institutions are dealing with vast quanitities of text. At Ulm University for instance, a machine-readable corpus of spoken language texts has been built up, with the purpose of support For psychotherapeutic process research. The corpus is administered by a Text Base Management System (TBMS), that integrates the functions of archiving, processing and analyzing an arbitrary amount of text (MERGENTHALER 1985). Several sysLmns satisfying the TBMS definition were conceived independently in the late seventies. THALLER (1983) reports a system CLIO, a TBMS with a highly differentiated data base component and a method base providing c~nputerized content analysis and comfortable editing. LDVLIB (DREWEK and ERNI 1982) is mainly a text analysis package, where data base management and text processing play a subordinate role. A PC-suited TBM-system, ARRAS (SMITH 1984), supports comfortable text inquiry by concordance and index functions, but has no textbase component. Finally there are two TBM-systems for commercial use, MIDOC (KOWARSKI and MICHAUX 1983) and MINDOK [INFODAS 1983) which have a database component and allow extensive processing of text, but no kind of text analysis at all. 2. Definition of a TBMS From the point of view of a TBMS-user, who is supposed to be a non-programming application-field worker, the system is an instrument to take up, to control and/or to analyze his or her individual texts for domain-dependent purposes. Consequently, a system intended to manage a text bank has the following tasks: 1. Input and editing of texts according to numerous points of view. 2. Management of an unlimited number of text units on a suitably sized auxiliary storage device. 3. Management of an unlimited amount of information concerning the text units, their authors, and the related text analyses. 4. Management of an open quantity of methods for editing and analyzing stored text units. 5. Assistance for interfaces to statistical and other user packages. 6. Assistance for a simple, dialogue-oriented user interface …