A Prediction Method for Destination Based on the Semantic Transfer Model

Qilong Han, Dan Lu, Kejia Zhang, Xiaojiang Du, Mohsen Guizani

Research output: Contribution to journalArticlepeer-review

5 Scopus citations

Abstract

With the widespread use of the mobile devices, destination prediction has become an important issue for location-based services (LBSs). Most existing methods are based on various Markov chain models, in which the predicted destinations are trained by historical trajectories. A problem among most of these follow-up works is that they blindly rely on the Markov process, ignoring the geographical distribution and the time property of the trajectories. In this paper, we propose a novel destination prediction algorithm, called STTL, based on the time property of the partial trajectory, along with the semantic transfer probability model trained in advance. We have conducted extensive experiments on the Shanghai Taxi dataset. The experimental results show that the STTL outperforms other state-of-the-art approaches.

Original languageEnglish
Article number8721085
Pages (from-to)73756-73763
Number of pages8
JournalIEEE Access
Volume7
DOIs
StatePublished - 2019

Keywords

  • Destination prediction
  • historical trajectories
  • semantic transfer probability
  • time-property

Fingerprint

Dive into the research topics of 'A Prediction Method for Destination Based on the Semantic Transfer Model'. Together they form a unique fingerprint.

Cite this