Distributed Anonymization for Multiple Data Providers in a Cloud System

Xiaofeng Ding, Qing Yu, Jiuyong Li, Jixue Liu, Hai Jin

Abstract: With the proliferation of cloud computing, there is an increasing need for sharing data repositories containing personal information across multiple distributed databases, and such data sharing is subject to different privacy constraints of multiple individuals. Most of the existing methods focus on single database anonymization, although the concept of distributed anonymization was discussed in some literatures, it only provides an uniform approach that exerts the same amount of preservation for all data providers, without catering for user’s specific privacy requirements. The consequence is that we may offer insufficient protection to a subset of people, while applying excessive privacy budget to the others. Motivated by this, we present a new distributed anonymization protocol based on the concept of personalized privacy preservation. Our technique performs a personalized anonymization to satisfy multiple data provider’s privacy requirements, and then publish their global anonymization view without any privacy breaches. Extensive experiments have been conducted to verify that our proposed protocol and anonymization method are efficient and effective.