Characterizing a user from large-scale smartphone-sensed data

Sha Zhao, Yifan Zhao, Zhe Zhao, Zhiling Luo, Runhe Huang, Shijian Li, Gang Pan

Abstract: Device analyzer can provide a large-scale dataset that captures real-world usage of smart phones [1]. Detailed usage records in smart phones, conveying a partial life log, are important for a deep scientific understanding of human characteristics. In this study, we proposed a feature-based labeling method to characterize users. Eight features from three aspects, i.e., daily mobility, user daily schedule, and social ability, are designed within a time window. Further, we analyze the features' correlation and variation over time. With the features, each user can be attached with a few semantic labels to demonstrate his/her characteristics. This work is a promising step towards drawing "portraits" for users using mobile phone data.