Using Answer Set Programming for Knowledge Representation and Reasoning: Future Directions

摘要

Since the proposal of the stable model semantics [1] of logic programs there has been a lot of developments that make answer set programs a suitable language for various kinds of knowledge representation. The building blocks that make answer set programming a suitable knowledge representation language include theoretical results, implementation and applications. The book [2] compiles most of the results that were available until 2002. Since then many additional results have been developed. However, many challenges and issues need to be further addressed before knowledge based intelligent systems become more prevalent.rnOne of the first challenge is to explore the extension of answer set programming so that it can model additional knowledge representation concepts in a natural way. Such concepts include representation of probabilistic information and representation and reasoning beyond the Herbrand Universe. Closely related to that is to explore ways to develop answer set programming systems that are not bogged down by grounding and can deal with numbers as easily as constraint logic programming systems do.