Autoepistemic logic of first order and its expressive power (Q1344888)

The autoepistemic reasoning process of rational agents in the autoepistemic logic of first order (first-order logic without modal operations, not a modal logic with quantifiers) is characterized by distinguishing positive and negative introspection. The maximal well- founded semantics for autoepistemic logic is defined on the two principles: positive introspection should be minimized while negative introspection should be maximized. Since the maximal well-founded semantics is based on full introspection, it defines a consistent set of theorems for any reasonable autoepistemic theories. Moore's AE logic and Reiter's default logic can be represented by autoepistemic logic. In the context of logic programming, the maximal well-founded semantics coincides with the preferential semantics. The author hopes that autoepistemic logic can be a good framework for nonmonotonic reasoning, logic programming, deductive databases and knowledge representation.