Managing Learner’s Affective States in Intelligent Tutoring Systems

Abstract

Recent works in Computer Science, Neurosciences, Education, and Psychology have shown that emotions play an important role in learning. Learner’s cognitive ability depends on his emotions. We will point out the role of emotions in learning, distinguishing the different types and models of emotions which have been considered until now. We will address an important issue concerning the different means to detect emotions and introduce recent approaches to measure brain activity using Electroencephalograms (EEG). Knowing the influence of emotional events on learning it becomes important to induce specific emotions so that the learner can be in a more adequate state for better learning or memorization. To this end, we will introduce the main components of an emotionally intelligent tutoring system able to recognize, interpret and influence learner’s emotions. We will talk about specific virtual agents that can influence learner’s emotions to motivate and encourage him and involve a more cooperative work, particularly in narrative learning environments. Pushing further this paradigm, we will present the advantages and perspectives of subliminal learning which intervenes without conscious perception. Finally, we conclude with new directions to emotional learning.