Trajectory Pattern Change Analysis And Applications In Wireless Networks

Mobile devices equipped with geo-location modules facilitateresearches on the nature of human mobility. In this paper, we aim to findeffective ways to mine user mobility patterns from the massive mobile data,analyze mobility pattern change, and use it for location prediction.While valuable contributions have been made to discover trajectorypatterns from movement history, there are still not a comparative amountof works specialized in analyzing how trajectory patterns change over time.In this paper, we facilitate this problem in a systematic and quantifiedmanner as identifying potential change points of user trajectory patterns.Specifically, we present a unified, information-based measure to quantifypattern changes between two intervals, and perform a Bayesian analysis ona series of aggregate measures to detect the actual change points.Experimenting on a three-month long dataset in real campus Wi-Finetworks, we show that our approach is effective to identify trajectorypattern changes in practice. Furthermore, we also inspect external factorsinfluencing user mobility in reality by associating trajectory patternchanges with public events, and there shows an interesting connectionbetween group pattern changes and real-world principles such as theweekday calendar or public events such as the May Day holiday.In addition, applying pattern analysis to location prediction, we findthat it can effectively help the prediction accuracy.