Modern society places many burdens on us and puts us at constant risk of mental health problems. Studies have shown that approximately one in three people have mental health problems, and counseling can be part of the solution. Counseling with people has been the norm, but it has been pointed out that this form of counseling has the psychological barrier of making people overly conscious of their therapist’s evaluation of them. Therefore, counseling with robots is attracting attention.
However, it is known to be difficult to establish a continuous relationship with a robot. This is because robots cannot accurately estimate a person’s condition and generate appropriate words and actions for each person. In order to accurately estimate a person’s state, it is necessary for the robot to extract features such as the facial expressions of the person being interacted with. In order for this to continue, the robot must continue to encourage self-disclosure to the person. Therefore, we believe that designing the robot’s words and actions to facilitate self-disclosure by the person will contribute to building a long-term relationship between the person and the robot and enhance the effectiveness of counseling.
Therefore, prior to the workshop, we implemented a dialogue system that encourages self-disclosure, and conducted an experiment in which participants interacted with a robot embedded with the system for one week. The workshop was then held for the participants of the experiment. During the workshop, participants were asked to think about how to establish a long-term relationship with the robot. The participants were also asked to think about where the robot could be applied to generate empathetic speech and behavior based on their estimated quality of life.