Improving the believability in the interaction of synthetic virtual agents: Towards Personality in Group Dynamics

Resumo

The interaction among humans and virtual agents should be provided for users like an xperience as natural as possible. Considering the same perspective of interaction found in humans, researchers define Personality as a key role for believable interface in group dynamics. In this paper, we describe how much Personality is fundamental to generate individuality in humans, and so, in virtual agents. By integrating those concepts in social agents we intend to conquer more believability during its interactions in groups of simulated societies. This paper presents a computational model that support the creation of group dynamics in teams of virtual agents that are engaged in cooperative tasks in order to create believability. Then, we describe an ongoing experiment based on a database of real agent.

KEYWORDS: Affective Computing, Personality, Social interaction in Lifelike Virtual Agents, Group dynamics.