9313061 Przymusinski This is the first year funding of a three-year continuing award. The main goal of the proposed research is to conduct a comprehensive study of non-monotonic formalisms and logic programming semantics for the purpose of developing and implementing formalizations that achieve greater expressive power by allowing various forms of disjunctive reasoning. Specifically, the following objectives will be achieved: 1) Study the basic principles on which disjunctive reasoning should be based and develop extensions of non-monotonic formalisms capable to properly represent disjunctive information. Investigate relationships between different approaches and determine their application domains. 2) Bridge the gap between non-monotonic reasoning and disjunctive logic programming by investigating and developing semantics for disjunctive logic programs and deductive databases that naturally correspond to non-monotonic formalisms. Attempt to develop proof-theoretic characterizations and computational techniques for these semantics and identify classes of theories in which they are tractable. 3) Develop methods of translation of broad classes of non-monotonic formalisms into disjunctive logic programs and implement an experimental inference engine for non- monotonic reasoning based on translation into disjunctive logic programs. 4) Apply the resulting techniques to the domain of software engineering by using disjunctive logic programming as a tool for software specification and verification of program correctness. If successful, the proposed research program should not only improve our understanding of basic foundations of non- monotonic reasoning, but it also should move us closer towards the actual implementation of common-sense reasoning systems by applying the increased expressive and computational power of disjunctive logic programming. It should also facilitate the extension of the logic programming paradigm to the class o disjunctive programs.
