earticle

영어

Recently, several cloaking methods based on K-anonymity and L-diversity has been proposed to protect the user’s location privacy for Location-based Services (LBS). Considering that a cloaking region could contain some semantic places, which can easily endanger the user’s privacy, it is not safe to cloak the user’s location only consider K-anonymity and L-diversity. This paper presents a novel personality privacy-preserving cloaking framework for the protection of sensitive positions on road-network environment. In our scheme, a Voronoi-partition graph is first learned from an urban network, and a Dominance Date Center (DDC) is introduced to take charge of the vertex’s Voronoi-partition (dominance space) data. Then, the θ-security semantics is introduced to measure the degree of sensitive semantics leakage. Thus, a lightweight agent running in the client can contract with DDC and process the sensitive semantics-aware cloaking algorithm to generate a cloaking region to meet K-anonymity and θ-security semantics. Final, not the anonymizer, but the client agent access direct into the LSP.