Obtaining user´s preference and ability from human-robot conversation towards mutual assistance

Community formation is important factor in human life, but it is not easy when your living environment is changed by retirement, moving, and so on. Therefore we focused on community formation and participation support system. Our proposed system find a relation and similarity by using personal characteristics. The authors place importance in a mutually assistance relation which utilize the advantages of each person´s ability, knowledge, and so on. In order to find a relation, the system have to obtain users´ personal characteristics, for example hobby, ability, preference, and so on. Therefore the authors are developing a dialogue robot system. Because the system have to provide a information or suitable relation according to individual preferences and interests that change from day to day. This paper shows that comparison human-human conversation and human-robot dialogue. The analysis result shows that conversation and dialogue include many information of personal characteristics. Moreover, the author tried one week free dialogue for long-term use. This trial also focused on time-dependence questionnaires. This result shows that timing of questionnaires are not make much difference of user´s answer, but time-dependence questionnaires might be encourage the dialogue.