The proposed research is based on an object-oriented semantic association model, which has been used in a prototype development of a distributed data administration system for supporting the integrated manufacturing facility of the National Bureau of Standards. This research aims to extend the model to allow user-defined operations and knowledge rules, which model integrity and security constraints, deductive rules, and domain specific expert rules, to be defined in object classes and instances. Thus, knowledge rules associated with objects are integrated with their data so that rules are readily available when associated data are being processed. A high-level object-oriented knowledge manipulation language will be developed. It allows the user to specify complex semantic patterns as search contexts and to concentrate on the objects of their interest without having to explicitly specify the traversal paths in queries. A mechanism for applying relevant rule-based knowledge while processing a transaction in a match-modify-execute cycle is proposed. The proposed functional integration of database and reasoning components allows tailoring of techniques such as query optimization, transaction management, integrity and security control, and concurrency control for use in the knowledge base management system environment.
