A Methodology to Design Human-Like Embodied Conversational Agents

Zacharie Alès (1) , Guillaume Dubuisson Duplessis (1) , Ovidiu Şerban (1) , Alexandre Pauchet (1)

1 LITIS - Laboratoire d'Informatique, de Traitement de l'Information et des Systèmes

This paper presents a bottom-up methodology to study human-human and human-agent dialogues in order to improve the design of embodied conversational agents (ECAs). The methodology uses a matrix representation of dialogues, constructed by means of a (semi-)automatic annotation process. A multidimensional dialogue pattern extraction and a clustering algorithm are applied to the coding matrices. The extracted patterns model the human behaviours.

Mots clés: Embodied conversational agents; a ffective and multimodal dialogue; automatic annotation; pattern extraction

Domaines: Intelligence artificielle [cs.AI]; Interface homme-machine [cs.HC]

Zacharie Alès, Guillaume Dubuisson Duplessis, Ovidiu Şerban, Alexandre Pauchet. A Methodology to Design Human-Like Embodied Conversational Agents. International Workshop on Human-Agent Interaction Design and Models (HAIDM'12), 2012, Valencia, Spain. online proceedings. ⟨hal-00927488⟩