Abstract
Fuzzy logic has been incorporated in many expert systems to solve real world problems that are inherently ambiguous. With fuzzy logic it is possible to program human intuition through the development of fuzzy expert system shells. A fuzzy expert system shell is a tool that helps build expert systems to manage fuzzy problems. Commercial as well as non-commercial fuzzy expert system shells are available. These shells provide variety of functions to facilitate the development of fuzzy expert systems for real world problems in different application areas such as medicine, engineering, and finance. To the best of our knowledge, none of the available fuzzy shells is natively developed for the Arabic language. This paper describes the development and the experimentation of a bilingual fuzzy expert system shell. This shell is intended to be a research tool for fuzzy expert systems developers in bilingual environments similar to those in the Arab world where users and developers use multi-languages due to their educational backgrounds and working environments. The shell processes fuzzy terms of the Arabic language as well as the English language. The shell is a general purpose shell that provides users with the ability to develop Arabic/English fuzzy expert systems using a simple Graphical User Interface. It applies implication methods that bear resemblance to human intuition. In the process of the development, a comparison of various fuzzy expert system shells has been performed to identify strengths and weaknesses of available shells. Experiments with our shell are reported and its performance is compared to existing shells that use different implication methods.