Defeasible Specifications in Action Theories

Chitta Baral and Jorge Lobo

Abstract

Recent research in formalizing effects of actions on a world in the presence of constraints have mostly concentrated on non-defeasible specifications, where the effect of actions and constraints are strictly specified. The focus of these research was more on solving the frame problem (and other problems such as the qualification and the ramification problem). In this paper we show how to incorporate defeasibility in the specification and at the same time solve the frame problem, and the other problems. In our approach we consider extensions of the high level language `A' of Gelfond and Lifschitz and introduce defeasible constraints and effect propositions. While direct semantics of `A' does not need a logical language, our semantics of the transition function (i.e., the effect of a single action) is defined using extended logic programming. This is due to the defeasibility of the specification. We then add additional logic programming rules to the logic programming characterization of the transition function to reason about the effect of a sequence of actions.