An intelligent consultant system for chess

Abstract

This paper deals with the problem of teaching chess. We describe the ideal characteristics for a learning environment for chess and discuss a partial implementation of such an environment; ICONCHESS. An Interactive Consultant for Chess Middlegames. Most research on computer chess has focused on creating highly competitive chess playing programs, regardless of the means used to achieve this goal. Because of the success obtained by programs based on search algorithms, little effort has been put on chess playing programs that play using high level strategies, which are necessary in human chess players. This lack of strategic foundations makes most chess playing programs inadequate as chess tutors too. This paper presents a new approach for computer chess in general and specifically for a learning environment for chess. This approach combines the techniques of fuzzy logic and case-based reasoning in order to produce high level advice for chess positions. We also present the results of the empirical experiments used to test ICONCHESS and suggest some ways in which the approach can be applied to other domains.