Privacy Protection in Secure Database Service

There has been much recent interest in secure database service that provides reliable storage and efficient query execution, while not knowing the contents of the database. Existing proposals for secure database services have typically been founded on encryption. However, our research shows that secure database services based on data encryption is not well-balanced on dealing with the contradiction between data privacy preserving and efficient queries. In this paper, we propose an automatic attributes detection partition method and a new security model which partition data in unencrypted form to distributed secure database servers. The theoretical analysis and experimental results show that our new method is feasible and provide efficient privacy protection and efficient query execution, and support horizontal fragmentation and semantic attribute decomposition.