Privacy preservation based on full-domain generalization for incremental data publishing

Abstract

As data can be continuously collected and grow all the time with the enabling of advancement in IT infrastructure, the privacy protection mechanism which is designed for static data might not be able to cope with this situation effectively. In this paper, we present an incremental full-domain generalization based on k-anonymity model for incremental data publishing scenario. First, the characteristics of incremental data publishing for two releases is to be observed. Subsequently, we generalize the observation for the multiple data release problem. Then, we propose an effective algorithm to preserve the privacy of incremental data publishing. From the experiment results, our proposed approach is highly efficient as well as its effectiveness, privacy protection, is very close to the bruteforce algorithm generating the optimal solutions.