An Emotion Generation Model Based on the Dialogist Likability for Sensitivity Communication Robot

Abstract

In recent years, robotics research has been shifting from industrial to domestic application, and several domestic, human centered robots, aimed at communicating expressively with human, have been developed. In this research, a novel robot, Ifbot, which can communicate with human by joyful conversation and emotional facial expression has been developed by our industry-university joint research project. For robots to live with people, communication, even if it is between robots and human, should involve not only conveying messages or instructions but also psychological interaction, such as comprehending mutual sentiments, sympathizing with the other person, and enjoying conversation itself. To communicate this way, a robot requires several mechanisms that enable the robot to recognize human emotions, to have emotions, and to express emotions. This paper focuses into an emotion generation method so as robot to have emotion. We think it is necessary that robot has a likability to it's dialogist for having humanlike emotion. This paper proposes an emotion generation model based on the dialogist likability and implement the model on Ifbot. This paper also reports the experiments of emotional conversation with Ifbot. We have confirmed from the experimental results that a change in emotion during conversation is caused by the difference of the dialogist likability, and the change brings a dialogist a certain psychological effect.