ABSTRACT
ABSTRACT: Since the location-based social networking services emerged, the location privacy issue has gained widespread concerns from consumers and researchers. While the location history plays an important role in knowledge discovery, many people worry that their privacy will be disclosed unsafely. In this paper, we introduce a spatial-temporal cloaking model which follows the k-anonymity principle to protect the location history. The corresponding algorithm which can find a minimum cloaking 3D box covering the actual location and its temporal information is proposed as well. Moreover, the algorithm is able to prevent the modeled records against the continuous attacks which infer the actual location by computing the overlapping spatial region. The experimental results show that the proposed algorithm has a high efficiency.
