Building Intelligent Emotion Awareness for Improving Collaborative E-Learning

In CSCL contexts it is necessary to study new forms of interaction that stimulate and promote the necessary skills for human communication and critical thinking in order to improve learning experience. Consequently, we have to take advantage of the interaction data generated in CSCL activities to extract important information related to the emotions manifested by group members and build emotional intelligence that supports awareness and regulation of emotions at both individual and group level. This raises important issues that concern, on the one hand, the ways to label affective behavior in educational discourse and, on the other hand, the ways to present emotional information to group participants. This paper focuses on the first issue and describes an approach that considers participant's affective behaviors when interacting in virtual spaces and labels their emotions in a non-intrusive way.