Access Control for Privacy Protection for Dynamic and Correlated Databases

The characteristics of volume, variety, velocity and value for big data have made present privacy protection methods less effective in the protection of user privacy to meet the emerging requirements. In this paper, we analyze the spatial and temporal effects the access to the data on privacy disclosure and propose an access control model to protect user privacy that is related to the number and the frequency of access in the access history of the requester. By introducing the notions of the privacy threshold, the requested items and the access history, the proposed access control model can make the decision on whether to allow the current access request to the protected privacy information. This method can cope with the dynamic and correlative nature of the data for privacy protection while reducing the cost of computation.