A Four-Participant Group Facilitation Framework for Conversational Robots

Abstract

In this paper, we propose a framework for conversational robots that facilitates four-participant groups. In three-participant con-versations, the minimum unit for multiparty conversations, social imbalance, in which a participant is left behind in the current conver-sation, sometimes occurs. In such scenarios, a conversational robot has the potential to facili-tate situations as the fourth participant. Conse-quently, we present model procedures for ob-taining conversational initiatives in incremen-tal steps to engage such four-participant con-versations. During the procedures, a facilitator must be aware of both the presence of dom-inant participants leading the current conver-sation and the status of any participant that is left behind. We model and optimize these situ-ations and procedures as a partially observable Markov decision process. The results of ex-periments conducted to evaluate the proposed procedures show evidence of their acceptabil-ity and feeling of groupness. 1