A Closer Look to Artificial Learning Environments

The so-called Intelligent Tutoring Systems evolved in the past five years to Intelligent Learning Environments. This shift was supported by the stronger role of software agents and the possible ways they can be used to promote human learning. The aim of this paper is to explore further on one possible style of formalization taking now a point of view from DAI in stead of AI and Education: a classroom is a small society of agents that can be simulated according to some specific goal defined by the user (say, the teacher). It will be based on a modification of the idea of deductive structures (DS), first proposed by Konolige. These DS will be accompanied by heuristic principles and criteria that define different types of societies and agents. The technical apparatus can be adopted in distinct educational scenarios from history to mathematics, and we will show, with an example, recent work in high school linear algebra.