FuzzRESS: A fuzzy rule-based expert system shell combining positive and negative knowledge for consultation of Vietnamese traditional medicine

Abstract

The aim of the paper is presenting an approach to developing a fuzzy rule based expert system shell combin-ing positive and negative knowledge for medical consultations called FuzzRESS. We extend Max-Min inference of CADIAG-2 like systems [3] by replacing Max of MaxMin rules by t-conorm and by including negative knowledge. Based on this approach, we propose a structure of FuzzRESS which consists of some main components: rule base, inference engine, explanation, interface and knowledge acquisition. The system FuzzRESS is implemented and demonstrated with diagnosis of Fever according to internal traditional medicine.