Abstract not found

For many commonsense reasoning tasks associated with action domains, only a relatively simple kind of causal knowledge is required - knowledge of the conditions under which facts are caused. This note introduces a modal nonmonotonic logic for representing causal knowledge of this kind, relates it to other nonmonotonic formalisms, and shows that a variety of causal theories of action can be expressed in it, including the recently proposed causal action theories of Lin. The new logic extends the causal theories formalism of McCain and Turner, and provides a more adequate semantic account of it. A useful subset of the logic has a concise translation into classical propositional logic, and so can be used for automated planning and reasoning about action. A larger subset is closely related to logic programming under the answer set semantics, yielding another approach to automated reasoning.

Most `causal' approaches to reasoning about action have not addressed the basic question of causality: what has to be the case in the world in order for the assertion `A causes B' to be valid? Pearl's recent causal theories based on structural equations do provide an answer to this question. In this paper, we extend Pearl's formalism so that the typical problems encountered in common sense reasoning about action can be represented in it. The resulting theory turns out to be a powerful tool for handling the ramification problem. It also provides new insights into actions with non-deterministic and/or `disjunctive ' effects and it may help bridge the conceptual gap between `causal' and `non-causal' approaches to common sense temporal reasoning. 1 Introduction Causality-based approaches to reasoning about action and change have attracted a lot of attention over the last few years [2, 13, 14, 15, 16, 17, 25], particularly in the context of modeling ramifications and actions with non-dete...