JPS-daprinfo: A Dataset for Japanese Dialog Act Analysis and People-related Information Detection

Abstract

Robots' or smart agents' memory construction and sharing have been proved to be important in human-robot interaction (HRI) and human-agent interaction (HCI) by many studies [1,2,3]. These studies suggest that regarding people-related information as a type of memory and sharing it with the user can be a good behavior for maintaining interaction. However, most of the relevant studies prepared the robot's memories manually, instead of obtaining them from robots' actual activities. This lack of memory constructing ability makes the strategy of sharing memories less practical.