Research Article
Predicting Next Points of Interests Based on a Markov Model
@INPROCEEDINGS{10.1007/978-3-030-00916-8_25, author={Jie Xu and Chunxiao Xing and Yong Zhang}, title={Predicting Next Points of Interests Based on a Markov Model}, proceedings={Collaborative Computing: Networking, Applications and Worksharing. 13th International Conference, CollaborateCom 2017, Edinburgh, UK, December 11--13, 2017, Proceedings}, proceedings_a={COLLABORATECOM}, year={2018}, month={9}, keywords={Next location prediction Markov chain model Cluster GPS trajectory analysis}, doi={10.1007/978-3-030-00916-8_25} }- Jie Xu
Chunxiao Xing
Yong Zhang
Year: 2018
Predicting Next Points of Interests Based on a Markov Model
COLLABORATECOM
Springer
DOI: 10.1007/978-3-030-00916-8_25
Abstract
With the development of Global Position System (GPS) technology, the analysis of history trajectory becomes more and more important. The Location Based Service (LBS) can provide the user’s location, the human movement location prediction from the history observations over some period have several potential applications and attract more and more attention. Predicting the user’s next position usually includes finding the Points of Interests (POIs) from the historical trajectory and predicting the position with a certain statistical model. In this paper, we present a novel method based on Markov chain for prediction, our method include two contributions: the first one we use GEPETO variant algorithm to cluster for POIs to solve the former algorithm without considering the temporal factor, and the second one we present Mobility Markov Chain (MMC) model which exploits 3 previous states to infer the future location. Our experiments basing on the real Beijing trajectories dataset display that our algorithm can improve the prediction accuracy compared with the baseline algorithm.