Modelisation d'un environnement informatique pour l'apprentissage humain a partir d'exemples base sur des principes d'auto-explication et de collaboration

Self-Explanation (SE) has been proved to be a useful activity specially for novices, while studying examples, yet ignored by many. In this master thesis, we designed, prototyped and tested a Computerized Assisted Learning Environment (CALE) based on a set of principles that we identified from literature on human learning by means of examples. The CALE that we propose, named SPEAC, exploits a domain-independent model of example. This model allows SPEAC to present examples in a pedagogical and structured way. A SE and collaboration module offers tools to help students self-explain and collaborate asynchronously for explaining different parts of an example. A first evaluation of SPEAC suggests that it encourages---at least allows---SE and impasse resolution through collaboration among peers, justifying the need for such an environment. An informal evaluation of SPEAC by its users also showed the importance of meta-informations in CALE exploiting new technologies of information and communication.