Mining spatial association rules from LBS anonymity dataset for improving utilization

Abstract

With the development of Location-Based Services system (LBS), privacy protection of LBS is becoming a hotspot topic in GIS and mobile communication domains. Among series of LBS users' privacies protection techniques, Spatial-Temporal K-Anonymity has become a prominent method for its easy implementation and extension. While, this method and its variants suffer from a common drawback that they decrease utilization of LBS anonymity datasets because of adopting the common principle that reduces spatial temporal resolution of LBS query. However, improving utilization of LBS anonymity datasets is very important for LBS providers. The reason is that it can benefit for many LBS applications. In this paper, we format basic concepts of mining spatial association rules from LBS anonymity datasets and design the implemented algorithms. In experiments we present the detailed process of mining spatial association rules which includes three phases: generating LBS anonymity datasets by adopting Spatial-Temporal K-Anonymity to GPS trajectories; preprocessing LBS anonymity datasets by spatial joining with geographic background GIS layers to achieve a spatial transaction database; mining spatial association rules from the spatial transaction database by adopting the proposed method. The experimental results show that utilization of the mined spatial association rules can assist intelligent traffic management.