(英) |
Heretofore, a subjective evaluation was used for evaluating the usability of a user interface (UI) such as a smartphone. However, the subjective evaluation does not accurately reflect the user's deep psychology. Therefore, in this research, we focus on brain activity and aim to estimate the pattern of user interface (UI) from the features of brain activity. The UI pattern of this research is a radio button and a check box (all unchecked). In this study, we analyze the data obtained by brain activity measurement and extract features quantity. As an estimation method, SVM (support vector machine) is used. Estimated model was constructed with closed data. And, Estimation accuracy was conducted with open data. As a result, for all subjects, high accuracy was obtained with closed data. Also, for specific subjects, high estimation accuracy of 0.8 or more was obtained with open data. In the case of the subjects with high discrimination accuracy of open data, was following. With the check box (all unchecked) of the UI pattern, it was found that the peak time appeared between 3 to 6 seconds. Also, the slope of the linear approximation tended to decrease. On the other hand, with the radio button of the UI pattern, it was found that there was no fixed trend in the peak time and the slope of the linear approximation. |