アスリートの発話に基づく心理状態の可視化アプリケーションの開発

抄録

This study proposed the analysis method for verbal data combining self-organizing maps and fuzzy cluster analysis in order to understand athletes’ mental states. The application for visualization of the results of analysis was also developed so that we can share the knowledge of the athletes’ mental states with non-psychological experts. The purpose of this study is to establish the method of psychological assessment using athletes’ verbal data through these process. The subjects of analysis were athletes’ utterances during interviews and their mental states were quantified in the form of vector data based on the utterances. These vector data (“psychological vector”) were mapped into two-dimensional plane (“map”) by means of self-organizing maps. The neurons composed of the map were clustered into ten fields and the seven kinds of the psychological evaluation axes were extracted from there. These results of analysis showed we can understand the mental change of each athlete by focusing on the transition of the psychological vector on the map. The developed application was expected to enable us to understand athletes’ mental states and changes more intuitively by expressing the transition using animation.