Developing Adaptive Interview Strategy Based on User's Attitude Recognition for a Interview Robot
ユーザーの態度推定に基づき適応的なインタビューを行うロボット対話システムの開発

Fuminori NAGASAWA, Takuya ISHIHARA, Shogo OKADA, Katsumi NITTA
JSAI Technical Report, SIG-SLUD  
The goal of this research is to develop a robot dialogue system that elicits story of users by choosing appropriate questions which user would like to answer. The key technique is the inference of users' motivation from multimodal information: gesture, posture, speaking status and prosodic status. In this paper, for this purpose, we developed a robot dialogue system that collects multimodal information of interviewees for examination of useful feature quantity for state estimation. Based on the
more » ... collected data, state estimation was performed by machine learning, and utterance motivation was estimated from multimodal information.
doi:10.11517/jsaislud.81.0_21 fatcat:grszubnfevgmhgbpokwch65o34