Fuzzy Object Match in Expert Systems

Fuzzy object, a new model of knowledge representation, is proposed and investigated. This model breaks through the traditional crisp models and presents a new approach in applying fuzzy theory and object oriented technologies to expert systems. The significance, knowledge representation structure and reasoning mechanism of fuzzy object model are analyzed in detail. Two methods of similarity degree and intersection function are used to calculate the matching degree between fuzzy objects and different fuzzy object patterns in rules. Finally, the possibility of realizing the idea of fuzzy object based on FuzzyCLIPS, a fuzzy expert system shell, is discussed.