Spatio-Temporal Trajectory Features Based Sensitive Relationship Protection

To solve the problem of privacy leakage of sensitive relationships caused by the spatial-temporal trajectory correlation of users in social networks, this paper proposes a privacy protection algorithm for sensitive relationships based on spatial-temporal trajectory features. In this paper, we propose a new measurement model for evaluating users' similarities, the basic idea of which is to calculate the similarity between users' sub-trajectories based on spatial and temporal dimensions. The proposed privacy protection algorithm adopts a heuristic to evaluate the inference contribution and information loss caused by data modification in order to protect sensitive relationship privacy meanwhile maintaining the trajectory data utility. We also provide the security analysis and theoretically prove the availability of the proposed algorithm. Based on the real social network data, the experimental results show that the proposed algorithm is efficient and could achieve high data utility.