Reference: Nayak, P. An Overview of Causal Reasoning. Knowledge Systems Laboratory, June, 1989.
Abstract: In recent years, causal reasoning has begun to play an increasingly important role in Artificial Intelligence. It has been used in a number of different application areas including medicine, digital circuits, and engineering systems. Causal reasoning systems stand in contrast to most current expert systems that are based on empirical associations. In most expert systems, the expert knowledge base typically consists of a set of rules that specify appropriate actions for a number of situations (called situation-action rules). These rules are usually culled from the experience of domain experts and reflect empirical associations between the situations and the appropriate actions. Causal reasoning systems take a different view to building knowledge bases. Instead of representing specific situation-action rules, causal reasoning systems represent the underlying causal knowledge in the domain. This causal knowledge is used for problem-solving and the specific situation- action rules are, in some sense, derivable from it. In this paper we provide an overview of the literature on causal reasoning.
Full paper available as hqx.