Subscribe to RSS - Affective Computing

Affective Computing

STIC-AMSUD Program | 2011 - 2012

Recent studies from psychology and neurology sciences have addressing the importance of emotions and other affective states for several intelligent cognitive processes, such as memory, decision taking and, inclusive, learning. These results have drawn the attention of Computers in Education researchers about the importance of taking into account students’ emotions in intelligent learning environments (ILE). In order to express emotions, the ILEs are generally incorporated with Embodied Conversational Agents (ECA). ECAs are computer-generated characters that are able to demonstrate many of the same properties as humans in face-to-face interaction, including the ability to produce and respond to verbal and nonverbal communication. As they are able to have a more anthropomorphic interaction with user, these agents have shown several benefits when included in ILEs, such as to engage students, focus their attention in important aspects of learning, demonstrate tasks, and so on. To be affective-aware, ECAs should have mechanisms in order to infer and express believable emotions. Although some results have already been acquired, real environments of interaction, as learning systems, require combining different methods and techniques with the purpose of obtaining more powerful inference mechanisms and generating more realistic verbal (text and speech) and non-verbal expression of emotions. This project aims at combining the Brazilian, Argentine and French investigation teams expertise in Artificial Intelligence applied to education and Affective Computing in order to create more affective-aware ECA and study its potential in real learning applications.