Learning Time-Varying Turn-Taking Behavior in Group Conversations

TL;DR

This paper introduces a probabilistic model that predicts turn-taking in group conversations by considering individual traits and past behavior, allowing for better generalization across different groups.

Contribution

It develops a generalized, data-driven probabilistic model that captures how individual speaking tendencies change over time in group conversations.

Findings

Model accurately predicts turn-taking in synthetic data.

Model effectively characterizes real-world group interactions.

Previous models may lack realism, highlighting the need for this approach.

Abstract

We propose a flexible probabilistic model for predicting turn-taking patterns in group conversations based solely on individual characteristics and past speaking behavior. Many models of conversation dynamics cannot yield insights that generalize beyond a single group. Moreover, past works often aim to characterize speaking behavior through a universal formulation that may not be suitable for all groups. We thus develop a generalization of prior conversation models that predicts speaking turns among individuals in any group based on their individual characteristics, that is, personality traits, and prior speaking behavior. Importantly, our approach provides the novel ability to learn how speaking inclination varies based on when individuals last spoke. We apply our model to synthetic and real-world conversation data to verify the proposed approach and characterize real group interactions. Our results demonstrate that previous behavioral models may not always be realistic, motivating our data-driven yet theoretically grounded approach.