Publication

The recent spread of smartphones derives growing interest in its application in health and medical care domains. Smartphone has the potential to reflect their owner's psychological states in sensor log and usage history. In this study, we propose a method to estimate the change in users' subjective well-being scales and those subcomponents using smartphone logs. For the precise estimation of well-being scales, the clustering algorithm is applied to separate similar user groups. The evaluation result revealed that binary classification accuracy (F1 score) was 0.899 by using a clustering algorithm and the change of seven subcomponents of the well-being scale was estimated between 0.682 and 0.874.