New method for building ASP knowledge base from knowledge in classical logic

Abstract

Answer Set Programming (ASP) is now a mainstream tool for the representation of non-monotonic knowledge. In order to make use of the existing knowledge in classic logic in the process of using ASP for problem solving, a method was proposed for translating knowledge in classic logic formulas to an ASP program or ASP knowledge base so that the models of the formulas and the answer sets of the ASP program were in one-to-one cor-respondence. Some examples were presented to illustrate the effectiveness of the method. Two classes of knowledge were distinguished in this paper, i.e. constraint knowledge that requires a formula to be satisfactory and definition knowledge that defines a predicate. In practice, the method provides a way of building non-monotonic ASP knowledge bases from the existing knowledge bases that use predicate logic as representation language.