Improving Skill for Self-Reviewing Presentation with Robot

Abstract

Presentation is very important for researchers to present their work. Proper presentation requires them to review it before actual presentation, which contributes to refining their presentation. In this work, we have focused on self-review, in which researchers review their presentation by themselves. A general way of self-review is to make a video of presentation, and then check it out. However, it brings about uncomfortable feeling due to looks and voice on the video, which prevents self-reviewing. We have accordingly developed a robot as presentation avatar, which reproduces presentation novel researchers (as learners) make. On the other hand, we have also confirmed that learners often overlook points to be modified in their nonverbal behavior reproduced by the robot. The main issue addressed in this paper is how to improve self-review skill with presentation robot. Our approach to this issue is to diagnose learners’ presentation by means of the presentation behavior model, and to help them compare their self-review results with the diagnosed results. Such comparison allows them to promote awareness of unsuitable/insufficient points in their self-review, which contributesto self-review skill development. In this paper, we demonstrate robot presentation comparison with two robots, in which one robot reconstructs their presentation with self-review results and the other one reconstructs it with diagnosed results. This paper also reports a case study with the system whose purpose was to ascertain whether robot presentation comparison contributes to promoting awareness of unsuitability/insufficiency in self-review. The results suggest the possibility of self-review skill development.