Titre : Modelling Cohesive Behaviors for Virtual Agents in Multiparty Interactions
Date : 8 octobre 2021
Résumé : Group interaction is a commonly used form of communication and humans communicate and coordinate with each other via a number of verbal and nonverbal behaviours. Group cohesion describes the members' attraction towards the group and the desire to be a part of the group. In this research work, we aim to recognize the relation between certain social cues and group cohesion. Further, researchers in artificial virtual agents community are integrating social behaviours to develop conversational agents. The main goal of this research is to develop a multiparty model involving multiple autonomous agents that can motivate and provide information to users. In particular, we are interested in modelling agents as a team capable of displaying cohesive group behaviour. In addition to this, since we are aiming to provide support and advice to the users it is also important for us to understand the effects of our model on the user and their perceptions e.g., the level of trust, persuasion.
Lieu : ISIR, 4 Place Jussieu, 75015, Paris