Host: The Japanese Society for Artificial Intelligence
Name : The 102th SIG-SLUD
Number : 102
Location : [in Japanese]
Date : November 28, 2024 - November 29, 2024
Pages 195-200
In this presentation, we propose a dialogue system that applies Motivational Interviewing (MI), a counseling technique aimed at promoting behavior change. MI is a dialogue method designed to elicit and strengthen a user's motivation for change. This system specifically aims to effectively elicit change talk, which are statements indicating a user's inclination towards change. The system operates in three stages: 1) classifying user utterances into talk types, 2) selecting the most appropriate response code based on the classification result, and 3) generating specific responses from the selected code. This approach enables appropriate interventions tailored to the user's state, facilitating effective dialogues aimed at behavior change. In this presentation, we will discuss the system's design details and report on preliminary evaluation results, as well as explore the potential and challenges of applying artificial intelligence to Motivational Interviewing.