AN IMPROVED ALGORITHM FOR INEXACT REASONING BASED ON EXTENDED FUZZY PRODUCTION RULES

Abstract

This paper proposes a new algorithm to deal with inexact reasoning, where extended fuzzy production rules are used for knowledge representation. Given the truth values of some conditions, the algorithm can perform inexact reasoning to evaluate the truth values of other conditions efficiently and automatically. First, the extended fuzzy production rules in a knowledge base are modeled by an augmented fuzzy rule matrix F. Then, based on the transitive closure of the augmented fuzzy rule matrix F, we develop an efficient algorithm for inexact reasoning. The algorithm proposed in this paper is better than the one presented in Chen et al. (1991) because it can handle the case of circular rules and can be executed more efficiently.