A fuzzy legal reasoner for university decision support

Abstract

Deciding on university student disciplinary cases against stipulated rules and case facts an interesting problem for artificial intelligence. Reasoning with fuzzy rules and facts adds to the problem complexity. This paper discusses the various characteristics of the problem and presents a design and an implementation of a prototype that is modeled as a fuzzy expert system. Some test results are presented and the experience gained from the project is discussed. Some future work is also suggested to further strengthen the prototype to include a formal case specification and interaction language, and the possible drawing and use of relevant information from a knowledge base of previous cases.