Research Focus
Mental Care Dialogue System Emphasizing Empathetic Active Listening
This study aims to develop a chatbot that supports daily mental health care by using a large language model, LLM. In modern society, many people experience stress and anxiety. Still, consultation with specialists is often delayed because of time constraints or psychological hesitation, such as wondering whether their concerns are serious enough to discuss. To address this issue, the proposed system incorporates knowledge from clinical psychology into prompts for the LLM. This design suppresses AI's tendency to provide immediate advice. Instead, it emphasizes active listening based on empathy and unconditional positive regard. In addition, the system leverages AI's anonymity and nonjudgmental nature to encourage users to disclose concerns they may otherwise hesitate to share with others. By providing an environment where users can easily express their feelings, this research aims to establish an autonomous mental care system that helps reduce psychological burden at an early stage.