Feedback-based reasoning process for behavior selection during long-term interaction

This paper focused on the effectiveness of the feedback-based reasoning process for managing long-term interaction. We propose the interaction process based on self-reference that a robotic system iteratively evaluates the status and the feedback during turn-based interaction. The result of reasoning process in the previous step is added to input queues, which also generates queries to reasoning to next behaviors. Assuming that reasoning process determines appropriate behaviors, we testified the effects of feedback-based reasoning in two different examples.