RECOGNITION AND GENERATION OF EMOTIONS IN AFFECTIVE e-LEARNING

This paper presents an educational system that incorporates two theories namely SAW and OCC in order to provide an improved affective e-learning environment. Simple additive Weighting (SAW) is used for the recognition of possible emotional states of the users, while the cognitive theory of emotions (OCC) is used for the generation of emotional states by educational agents. The system bases its inferences about users' emotions on user input evidence from the keyboard and the microphone, as two commonly used modalities of human-computer interaction. The actual combination of evidence from these two modes of interaction has been performed based on a sophisticated inference mechanism for emotions and a multi-attribute decision making theory. At the same time, user action evidence from the two modes of interaction also activates the cognitive mechanisms of the underlying OCC model that proposes emotional behavioural tactics for educational agents who act for pedagogic purposes. The presented educational system provides the important facility to authors to develop tutoring systems that incorporate emotional agents who can be parameterized so as to reflect their vision of teaching behaviour.