A Rough Set Based Approach for Solving the Unmatching Problem between User Query and Rules in Expert System

摘要

This paper puts forward an approach to solve the unmatching problem between user query and rules for expert system. The approach partially matches user query with rules, and then select the nearest rule for the query. Both of the weight of rules and the significance of condition attributes are considered. Weights of rules are decided by support and confidence of rules while the significance of each condition attribute is objectively calculated on the basis of rough set theory. The right hand side of the nearest rule is the solution for user query. Feasibility of the approach is demonstrated by a simple example.