Encoding probabilistic causal models in probabilistic action language PAL.

Chitta Baral and Nam Tran

Abstract

Pearl's probabilistic causal model has been used in many domains to reason about causality. Pearl's treatment of actions is very different from the way actions are represented (explicitly) and their impact is reasoned in most other papers in the literature. In this paper we show how to encode Pearl's probabilistic causal model in the action language PAL thus relating this two distinct approaches to reason about actions.