Intelligent Tutoring System with Affective Behavior

We have developed an affective behavior model for intelligent tutoring systems. The aim of the model is to provide students with a tutorial action taking into account both, the affective and the pedagogical state of the student. The affective behavior model considers the student’s pedagogical and affective state, this last one is based on the OCC and Five-factor models. To select the tutorial actions according to the pedagogical and affective state, we propose the use of a dynamic decision network with a utility measure on both, learning and affect. We have integrated the affective behavior model to an educational game to learn number factorization and evaluated its performance with a group of students. To evaluate the affective behavior model we designed a user study with a group of students playing with the educational game without the affective behavior model, and another group that included the affective behavior model. After that, we compared the learning gain in both groups. In this paper we describe the affective behavior model and present the results of the user study.