MCFS: a multiple criteria reasoning fuzzy expert systems building tool

Abstract

The authors present the design principles of MCFS, an expert system building tool based on the idea of combining multiple criteria reasoning with the concepts of fuzzy logic. An important feature of MCFS is its ability to handle multiple criteria reasoning by structuring the deduction process into a hierarchy of logical levels. Within each level, the rules are organized into sets of rules and rule set groups. The concepts of fuzzy logic are introduced by allowing uncertainty within each rule, each set of rules, and each rule set group. MCFS has been fully implemented and applied to several domains of knowledge such as computer system selection and procurement, solution of nonlinear simultaneous equations, neck-tie selection, and longevity estimation. Experiments with these applications indicate that, compared to standard expert system tools, MCFS produces expert systems which are easier to build and better match the human expert.<>