Evaluation of the teaching-learning process with fuzzy cognitive maps

Abstract

The evaluation of the teaching-learning process is a conduct that poses great challenges from a modeling perspective. Tutor modules form part of intelligent systems that are applied to education, and should assess: when to interrupt, how to teach a given topic, and what to teach in a given moment. By considering the different types of errors, a good evaluation of the teaching-learning process allows the systems to adapt in an optimum form to different users. This paper proposes a design technique to model the conduct of an expert that evaluates the results of the teaching-learning process within a reactive learning environment. To achieve this, a fuzzy cognitive map, which is a representation recently proposed to model the conduct and operation of complex systems, is used. Bart Kosko introduced this type of maps in 1986 with the aim of describing the behavior of a system in terms of concepts and causal relations between concepts. Since then, their use has been extended to diverse real-world situations that span from the analysis of investment in stock to supervisory system control. Missing conceptions, which provide a basis to assemble the didactic tactics, are imputed to this representation. Starting from the mental model of the expert in the learning domain, a conceptual genetic graph is established to orientate the fuzzy cognitive map. Finally, this paper presents a concrete example. © Springer-Verlag Berlin Heidelberg 2004.