Knowledge-Based System in an Affective and Intelligent Tutoring System

Abstract

This book chapter presents an affective and intelligent tutoring system called Fermat that integrates emotion or affective states with an Intelligent Learning Environment. The system applies Knowledge Space Theory to implement the knowledge representation in the domain and student modules and Fuzzy Logic to implement a new knowledge tracing algorithm, which is used to track student’s pedagogical and affective states. The Intelligent Learning Environment was implemented with two main components: an affective and intelligent tutoring system for elementary mathematics and an educational social network. The tutoring system generates math exercises by using a fuzzy system that is fed with cognitive and effective values. Emotion recognition was implemented by two methods: one for feature extraction of the face and one for feature classification using back-propagation neural networks. In addition to recognizing the emotional state of the user, our system gives emotional support through a pedagogical agent. Furthermore, an architecture of software is presented where the emotion recognizer collaborates with the affective and intelligent tutoring system inside a social network. Finally, we present a real-time evaluation with third-grade students in two different schools.