Exploring Home and Work Locations in a City from Mobile Phone Data

Recently, mobile phone call detail records (CDR) have been used to study mobility patterns in cities. Since home and workplace are the most important places in people's lives and define the structure and activity patterns of the city, identifying home and work locations and home-work commuting patterns are of much interest. However, due to decreasing usage of voice calls and SMS, and low usage while people staying at home, identifying home and work locations are getting more difficult. In this paper, we develop a method to exploit daily-aggregated internet usage data (G-CDR) in addition to CDR for identifying work locations. We also develop a method for identifying home locations based on detecting sleeping time. This method can significantly increase home detection success rate and accuracy. From the identified home and work locations, we have explored the population distribution and commuting patterns of people in Bangkok.