AN OBJECT-ORIENTED DATA MODEL FOR ARTIFICIAL-INTELLIGENCE APPLICATIONS

The aim of this paper is to show that a wide class of AI applications can be directly supported by Object-Oriented database technology. The class of AI applications addressed is that of Terminological systems, such as KL-One and Back, characterized by Frame Definition Languages. In this paper we present the Object-Oriented data language OOL with the intention of assessing its adequacy in supporting AI applications. To this end, we compare OOL with FL, a typical knowledge representation language designed for AI applications. OOL allows the definition of types, similar to concepts in FL. In the first part of the paper, the syntax and the formal semantics of OOL are introduced, then the rules to compute subsumption between types are defined. Finally, it is shown that subsumption for OOL is sound and complete, furthermore we prove that OOL is tractable though it results to be more expressive than FL-, the tractable form of FL.