Analysis of dialogue for acquiring personal characteristics toward co-occurrence matching

摘要

Eldely person living alone is easy to stand alone from the society. When a desaster happens, it is difficult to help them instantly. After 3.11, the importance of local community has been recognized. So, we form a community including the elderly person living alone in daily life. We should promote the cooperation not only daily life but also emergency. To form a community, it is necessary to make human relationship. There are the information recommendation systems by using cooperation filtering method. However, there is a burden on users during inputting data and we aren't able to find the relation of co-occurrence to help each other. In previous study, the authors conducted a experiment to acquire personal characteristics by human-robot conversation. And we analyzed results of experiment and matching by manual. Therefore in this research, we conducted a experiment to acquire personal characteristics by human-robot conversation including daily conversation contents. Moreover, we propose the method to value personal characteristics by numerical value and to do matching based on personal characteristics.