Using human memory structures to model knowledge within algebra virtual laboratory

In this paper we describe our knowledge representation model embodied within an algebra virtual laboratory dedicated to Boolean reduction problem solving. The model is inspired by psychology theories that explain the human cognitive activity in terms of memory subsystems and their processes. During the problem solving, complex knowledge entities, built from primitive units of knowledge, are dynamically combined to represent the learner behavior. This has the advantage to offer a closely fine prediction of the mastery degree and the acquisition level of each element of the taught knowledge.