Discovering Individual Life Style From Anonymized WiFi Scan Lists on Smartphones

Abstract

The prevalence of smartphones equipped with various sensors enables the pervasive capture of users' location data. WiFi scan lists on one smartphone, i.e., scan results of the network in a range, can roughly indicate the physical location of the phone in a time period. Considering the close relationship between location and daily life, users' lifestyle can be inferred from their WiFi scan lists. Given the issue of user privacy, in this paper, we explore anonymized WiFi scan lists to discover users' lifestyle. Individual lifestyle about mobility and important places of home and workplaces is discovered, respectively, based on the stay places extracted from anonymized WiFi scan lists and the reconstructed mobility trajectories. We first learn the lifestyle about mobility by detecting activity areas from mobility trajectories and introducing two metrics of activeness and diversity to measure individual mobility. Then, we discover the lifestyle about the home and workplaces identified from anonymized WiFi scan lists, such as stay duration at home, activeness of going outside at night, and working hours on weekdays and weekends. The experiments were conducted on a real-world large-scale dataset, which contains records of smartphone usage of more than 17 000 volunteering participants. Our work is a promising step toward automatically discover people's lifestyle from anonymized smartphone data.