The Impact of Temporal Factors on Mobility Patterns

Many applications such as management of wireless networks and the spread of mobile or biological viruses depend on modeling and predicting human mobility. However, widespread wireless localization technology, such as pervasive cell-tower/GPS location estimation, has only been available for the last few years, thus many factors that impact human mobility patterns remain under-researched. In this paper, we investigate how temporal factors impact mobility characteristics and location prediction. Our analysis of 180 mobile phone location traces show that people move farther distances, choose more unpredictable locations to visit, and have a more scattered spatial probability distribution for their location on the weekends compared to week days, or at after-work hours compared to work hours. We also analyzed location traces for a month and divided days and hours into groups for each user to obtain probability distribution of their places for each group of time intervals, and observed major improvement in future 'time-based' predictions of their location.