2015 IEEE International Conference on Data Science and Data Intensive Systems (DSDIS) (2015)
Dec. 11, 2015 to Dec. 13, 2015
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/DSDIS.2015.88
Location prediction based on cellular network traces is a very challenging task due to the randomness of the human mobility patterns. With the help of the abundant social interaction data contained in the cellular network, this paper focus on this question: How can knowing the location and the assembled and dismissed behavior of my friends be used to more accurately predict my location? We find out that the collective effect users' mobility spontaneously. We tested it in 2 ways. And find that 1: User tend to remain himself at the area where his friends are denser. 2: The more impact from friends would come from my more favorite place. And we present a prediction model according to these two phenomena. The result shows that this model did improve location prediction from 2.6% to 12.9%, for the users with enough social information.
Predictive models, Frequency measurement, Mobile handsets, Context, Global Positioning System, Association rules, Silicon
C. Zhou, B. Huang and L. Tu, "Exploiting Collective Spontaneous Mobility to Improve Location Prediction of Mobile Phone Users," 2015 IEEE International Conference on Data Science and Data Intensive Systems (DSDIS), Sydney, Australia, 2015, pp. 117-122.