Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Let's move on: Topic Change in Robot-Facilitated Group Discussions

Georgios Hadjiantonis, Sarah Gillet, Marynel Vázquez, Iolanda Leite, Fethiye Irmak Dogan

TL;DR

This work tackles autonomous topic-change decisions in robot-moderated group discussions by casting the problem as a three-class classification over multimodal cues. It compares non-sequential and sequential ML models, plus two baselines, on a newly collected Shutter robot dataset annotated with not appropriate, appropriate, and needed labels, using a 2s to 2s context around utterances. Key findings show that learning-based approaches generally outperform handcrafted heuristics, with acoustic features providing robust predictive power and two-step binary classification offering strong performance for not appropriate vs appropriate/needed. The study demonstrates the feasibility of real-time, content-free topic management in robot-facilitated groups and provides a publicly available dataset to support further research in autonomous HRI topic moderation.

Abstract

Robot-moderated group discussions have the potential to facilitate engaging and productive interactions among human participants. Previous work on topic management in conversational agents has predominantly focused on human engagement and topic personalization, with the agent having an active role in the discussion. Also, studies have shown the usefulness of including robots in groups, yet further exploration is still needed for robots to learn when to change the topic while facilitating discussions. Accordingly, our work investigates the suitability of machine-learning models and audiovisual non-verbal features in predicting appropriate topic changes. We utilized interactions between a robot moderator and human participants, which we annotated and used for extracting acoustic and body language-related features. We provide a detailed analysis of the performance of machine learning approaches using sequential and non-sequential data with different sets of features. The results indicate promising performance in classifying inappropriate topic changes, outperforming rule-based approaches. Additionally, acoustic features exhibited comparable performance and robustness compared to the complete set of multimodal features. Our annotated data is publicly available at https://github.com/ghadj/topic-change-robot-discussions-data-2024.

Let's move on: Topic Change in Robot-Facilitated Group Discussions

TL;DR

This work tackles autonomous topic-change decisions in robot-moderated group discussions by casting the problem as a three-class classification over multimodal cues. It compares non-sequential and sequential ML models, plus two baselines, on a newly collected Shutter robot dataset annotated with not appropriate, appropriate, and needed labels, using a 2s to 2s context around utterances. Key findings show that learning-based approaches generally outperform handcrafted heuristics, with acoustic features providing robust predictive power and two-step binary classification offering strong performance for not appropriate vs appropriate/needed. The study demonstrates the feasibility of real-time, content-free topic management in robot-facilitated groups and provides a publicly available dataset to support further research in autonomous HRI topic moderation.

Abstract

Robot-moderated group discussions have the potential to facilitate engaging and productive interactions among human participants. Previous work on topic management in conversational agents has predominantly focused on human engagement and topic personalization, with the agent having an active role in the discussion. Also, studies have shown the usefulness of including robots in groups, yet further exploration is still needed for robots to learn when to change the topic while facilitating discussions. Accordingly, our work investigates the suitability of machine-learning models and audiovisual non-verbal features in predicting appropriate topic changes. We utilized interactions between a robot moderator and human participants, which we annotated and used for extracting acoustic and body language-related features. We provide a detailed analysis of the performance of machine learning approaches using sequential and non-sequential data with different sets of features. The results indicate promising performance in classifying inappropriate topic changes, outperforming rule-based approaches. Additionally, acoustic features exhibited comparable performance and robustness compared to the complete set of multimodal features. Our annotated data is publicly available at https://github.com/ghadj/topic-change-robot-discussions-data-2024.

Paper Structure

This paper contains 24 sections, 1 equation, 2 figures, 5 tables.

Table of Contents

  1. INTRODUCTION
  2. Related work
  3. Dataset
  4. Annotations
  5. Feature Extraction
  6. METHODOLOGY
  7. Problem Formulation
  8. Non-sequential Data Modeling Approach
  9. Sequential Data Modeling Approach
  10. Baselines
  11. Feature-Based Heuristic
  12. Speech-and-Pause-Based (SPB) Heuristic
  13. Experiments
  14. Evaluation Procedure
  15. Data Splitting and Balancing
  16. ...and 9 more sections

Figures (2)

  • Figure 1: The interaction between the robot and three participants from two different perspectives. The robot moderates the group discussion and needs to decide when to move on to the next topic.
  • Figure 2: Utterance sequence from our annotated dataset. The interaction is between the robot (R) and two participants (P1 and P2), discussing the question: "Is there any way that a robot like me could be helpful in those places in your home you just described?".