The study of belief change has been an active area in philosophy and AI. Inrecent years two special cases of belief change, belief revision and beliefupdate, have been studied in detail. In a companion paper, we introduce a newframework to model belief change. This framework combines temporal andepistemic modalities with a notion of plausibility, allowing us to examine thechange of beliefs over time. In this paper, we show how belief revision andbelief update can be captured in our framework. This allows us to compare theassumptions made by each method, and to better understand the principlesunderlying them. In particular, it shows that Katsuno and Mendelzon's notion ofbelief update depends on several strong assumptions that may limit itsapplicability in artificial intelligence. Finally, our analysis allow us toidentify a notion of minimal change that underlies a broad range of beliefchange operations including revision and update.
展开▼
