Proceedings Abstracts of the Twenty-Third International Joint Conference on Artificial Intelligence

Functional Stable Model Semantics and Answer Set Programming Modulo Theories / 718
Michael Bartholomew, Joohyung Lee

Recently there has been an increasing interest in incorporating "intensional" functions in answer set programming. Intensional functions are those whose values can be described by other functions and predicates, rather than being pre-defined as in the standard answer set programming. We demonstrate that the functional stable model semantics plays an important role in the framework of "Answer Set Programming Modulo Theories (ASPMT)" — a tight integration of answer set programming and satisfiability modulo theories, under which existing integration approaches can be viewed as special cases where the role of functions is limited. We show that "tight" ASPMT programs can be translated into SMT instances, which is similar to the known relationship between ASP and SAT.