[3P5-OS-9a] OS-9

[3P5-OS-9a-04] Encouraging Self-Disclosure through Adaptive Interview Dialogue Based on Multimodal Attitude Estimation

〇Fuminori Nagasawa¹, Shogo Okada¹ (1. Japan Advanced Institute of Science and Technology)

Keywords:Dialog System, Multimodal Sentiment Analysis

The goal of this research is to develop a communicative robot as a partner that can share what users want to talk about and what they are interested in.
To achieve this goal, this study proposes an interviewer robot that adapts topics according to the user's multimodal attitudes. The robot estimates whether a topic change or continuation is appropriate based on the multimodal features of the user during the dialogue, and generates questions based on the estimated results using a large-scale language model. This paper reports the results of an evaluation of the impact of an adaptive interview dialogue strategy on user self-disclosure.
The results of an experiment with 30 subjects showed that when the topic was changed adaptively, the degree of self-disclosure was significantly higher than when the topic was changed randomly. A correlation was found between the correct response rate for various attitude estimations and the degree of self-disclosure, indicating that self-disclosure becomes more active when the accuracy of attitude estimation improves.