Required Precision to Natural Language Processing for Therapeutic Patient-Health Care Provider Communication

K. Osaka, K. Matumoto, C. Kawanishi, T. Tanioka, T. Tada, S. Ueno (Japan), R.C. Locsin (USA), and F. Ren (Japan, PRC)


natural language processing, therapeutic communication, human support robot


It is necessary to approach the caring humans perform by achieving empathetic understanding in human-robot communication and expressing caring feelings for a human-supporting robot. This paper aims at requiring precision to natural language processing for therapeutic patient-health care provider communication. First, our subjects were asked to view a video, which was to induce emotions. And then, an interview was carried out in order to identify the scenes that caused changes in the subjects’ emotion. Also, the scenes were objectively judged by measuring the activity of an electroencephalogram (EEG) of the subjects. Finally, the lines of the scenes were extracted and, taking the subjects’ impressions of their empathy as a criterion for judgement of natural language processing (NLP), compared with the analysis results by NLP. By our experiment, although it was difficult to make a judgment on the impressions in which emotions were intermingled such as joy and sorrow, NLP and the criterion for judgement that was the empathetic understanding of the subjects were consistent. It is clear that an empathetic understanding performed in medical welfare sites should be able to sensitively perceive pain, distress and loneliness of the subjects in particular.

Important Links:

Go Back