AI inferences utilizing Occam Abduction

Abstract

Abduction is formally defined as finding the best explanation for a set of observations, or inferring cause from effect. Here we discuss the notion of Occam Abduction, which relates to finding the simplest explanation with respect to inferring cause from effect. Occam abduction is useful in artificial intelligence in application of autonomous reasoning, knowledge assimilation, belief revision, and works well within a multi-agent AI framework. Here we present a flexible, hypothesis-driven methodology for Occam Abduction within a cognitive, artificially intelligent, system architecture.