HIB-tree: An efficient index method for the big data analytics of large-scale human activity trajectories

摘要

Research on traditional trajectory data has been widely carried out, and the popularization of mobile social network and location based service have enabled traditional trajectory data correlate with more human activity information, named as activity trajectory. How to generate large-scale, high-quality activity trajectories and provide efficient indexing structure to support big data management and analysis, has become a new challenge. This paper proposes a new method to generate activity trajectory, which produces relevant keywords for activity trajectory based on word embedding, then an efficient index HIB-tree and an enhanced similarity algorithm based on point popularity and keyword influence factor are proposed to recommend activity trajectory. Experiments show that the performance of HIB-tree is better than the existing index methods of activity trajectory.