Improvement of User’s Attitude toward Decision-Making Task by Multimodal Implicit Expressions of Agent’s Subjective Attitude

Abstract

It is expected to realize an agent that supports user’s decision-making and enhances user’s intellectual activities through interaction. On the other hand, such support may reduce the user’s individuality and development. We considered the possibility that an agent who supports decision-making while interacting with a user can improve understanding, impressions, and attitudes toward decisions without interfering with the user’s individuality by conveying the agent’s subjective attitude through implicit multimodal expressions. To investigate the hypothesis, we implemented an implicit-expression agent and an explicit-expression agent and conducted an experiment. The results showed that the implicit-expression agent was rated higher than the explicit-expression agent. Despite the explicit-expression agent having an advantage in conveying the agent’s intentions, these results suggest that implicit expressions positively affect the decision-making task.