GAITS II: An intelligent system for computer-aided education

Abstract

The design of computer-aided education systems is an important area that offers many challenges to the artificial intelligence community. The problems involved in such design are ambiguous and the managed data are imprecise. The coupling of symbolic and numerical computing might be the key for making educational environments more robust, and better in handling the involved problems. This paper presents mechanisms for coupling symbolic and numerical computing in intelligent educational environment. The proposed method combines heuristic AI methods with analytical control of adapting computer-aided education. A coupled system (GAITS II) with symbolic and numerical frameworks is described in this paper. A rule-based inference engine for diagnostic reasoning is coupled with an optimization method to select an appropriate dialogue (prototype remediation) for a student. The system deals with both the symbolic and numerical methods using not only an inference engine but also fuzzy sets theory and optimization methods (genetic algorithms). We present a new optimization method to select an appropriate dialogue for a student defining a new global mapping function between the pedagogical guidelines and the corresponding numerical terms.