A Parallel Clustering Algorithm Based on Grid Index for Spatio-temporal Trajectories

The clustering process in pattern mining of spatio-temporal trajectories is an important research content. Although there has existed extensive research on trajectory clustering, the efficiency of these algorithms is not able to meet the efficiency requirements faced with the large volumes of position data from moving objects. In order to improve the efficiency of clustering, this paper proposes two parallel algorithms for trajectory clustering, using parallel computing to seed up the calculation and employing the grid index to realize regional query and reduce some unnecessary calculations. Finally, the efficiency of the proposed algorithms is validated by extensive experiments on a real taxi trajectory dataset.