On Context-based Naming in Information Bases

Manos Theodorakis and Panos Constantopoulos
Department of Computer Science , University of Crete
and
Institute of Computer Science , Foundation for Research and Technology - Hellas
P.O.Box 1385, GR 711 10 Heraklion, Crete, Greece
email: { etheodor | panos }@ics.forth.gr

Abstract:

In information bases following semantic and object-oriented data models logical names are used for the external identification of objects. Yet the naming schemes employed are not ``natural'' enough and several problems often arise: logical names can be ambiguous, excessively long, unrelated to or unable to follow the changes of the environment of named object. In natural language, similar problems are resolved by the context within which words are used. An approach to introducing a notion of context in an information base is to provide structuring mechanisms for decomposing it into possibly overlapping parts. This paper focuses on developing a context mechanism for an information base and, in particular, exploiting this mechanism for naming purposes. Rules are developed for generating meaningful names for objects by taking their context into account. This context-based naming enhances name readability, resolves name ambiguities, saves a lot of redundant name substrings, and it localizes and thus facilitates consistency checking, query processing and update operations. In modeling, it supports systematic naming of objects, and thus enhances cooperation between the designers and the end-users in the sense that the contents of the information base are more understandable by both of them.

Manos Theodorakis
Tue May 14 12:47:46 EET DST 1997