An Expert System Framework for Non-Monotonic Reasoning about Probabilistic Assumptions

Abstract

Publisher Summary This chapter discusses an expert system framework for nonmonotonic reasoning about probabilistic assumptions. A major concern in current research on expert systems is the incorporation of methods for representing and manipulating uncertainty. Attempts to replicate probabilistic reasoning in expert systems, however, have typically overlooked a critical ingredient of that process. The present work addresses this problem in the context of conflict resolution in expert systems for image analysis. It involves the design of an expert system inference framework in which probabilistic statements are regarded as assumptions, which are explicitly tracked and reevaluated when they lead to conflict among different sources of evidence or lines of reasoning. Two conceptions of conflict and conflict resolution have been implicit in most approaches to this area. From one point of view, divergence among lines of reasoning can be regarded as stochastic; it is expected to occur some small percentage of the time, even when both processes of reasoning are normatively optimal, because of the imperfect correlation between cues and hypotheses or the chance accumulation of small errors in measurement.