33 Received: September 17th, 2007 Revised: September 30th, 2007 ABSTRACT. In this paper, we discuss the social intelligence that renders affective behaviours of intelligent agents, and its application to a collaborative learning system. We argue that socially appropriate affective behaviours would provide a new dimension for collaborative learning systems. The description of a system to recognize the six universal facial expressions (happiness, sadness, anger, fear, surprise, and disgust) using an agent-based approach is presented. Then, we describe how emotions can efficiently and effectively be visualized in Collaborative Virtual Environments, or CVEs, with an animated virtual head (Emotional Embodied Conversational Agent, or EECA) that is designed to express and act in response to the “universal facial expressions”. The objective of the paper is to present the emotional framework EMASPEL, that is Emotional Multi-Agents System for Peer to peer E-Learning, based on the Multi-Agents Architecture approach. KEYWORDS: Agent-based approach, Collaborative learning systems, Collaborative virtual environments, Emotional agents, Peer-to-peer e-learning Emotional agents for collaborative e-learning Mohamed Ben Ammar, University of Sfax, Sfax,Tunisia Mahmoud Neji, University of Sfax, Sfax, Tunisia Adel M. Alimi, Université de Pau et des Pays de l’Adour, France Guy Gouardères, Université de Pau et des Pays de l’Adour, France Problem formulation The field of affective computing was proposed and pioneered by Rosalind Picard from the MIT Media Laboratory. Her definition of affective computing is: “computing that relates to, arises from, or deliberately influences emotions” (Picard, 1997). Her argument for putting emotions or the ability to recognize emotions into machines is that neurological studies have indicated that emotions play an important role in our decision-making process. Our “gut feelings” influence our decisions. Fear helps us survive and avoid dangerous situations. When we succeed, a feeling of pride might encourage us to keep on going, and push ourselves even harder to reach even greater goals. Putting emotions into machines makes them more human, and should improve human-computer communication. Also exploiting emotions could lead to a more human decision-making process. Consequently, in this paper, the emotional agents for collaborative e-learning Ammar - Neji - Alimi - Gouardères