On predicting the residence time of mobile users at relevant places

Predicting future spatial and temporal behavior of mobile users is essential for the development of a wealth of new applications and services. In this paper, we focus on predicting the residence time of a user at their relevant locations. We explore the joint use of location history, arrival time, and the previous residence time to accurately predict the residence time at the current location. We developed a model that integrates all these parameters and uses our modified k-moving-average and k-CDF time-aided algorithms that include the arrival time in the model. We run performance evaluation experiments on a large real mobility trace collected by Dartmouth College and made publicly available through the CRAWDAD project. The dataset we worked on included 545 access points and 6.181 users. Our results show that adding high-granularity temporal information to the mobility model allows to significantly improve the residence time prediction compared to state-of-the-art methods. The prediction accuracy improvement for the dataset we work on has been consistent and of about 20% on the average.