Approaches For Privacy Preserving Based On Individual Correlation In Continuous Data Release

Continuous data release,which is a kind of data release,means that the data owner releases the updated data with the same structure continuously over time.Due to the rapid development of modern social network,the previous data is easy to be collected,and the individual correlation in data can easily be obtained from various network channels by attacker.Based on this background knowledge the attackers can easy get personal privacy information.Therefore,it is of great significance to study and improve the privacy preserving methods to resist the new attacks in the continuous data release.Almost no existing privacy protection method has ever take the attack caused by the individual correlation into account in continuous data release.Aiming at this problem,in this paper,starting with analysis of the attack based on individual correlation in continuous data release scenario,we design a privacy protection model that can resist this attack,namely privacy protection model of continuous data release based on individual correlation.In the privacy protection model,we first design a property that correlation tuples' anonymous group shoud satisfy,called ? group senstive values invariance.Then we give different conditions that different kind of tuples' anonymous group shoud satisfy and finally theoretically demonstrated its safety.Based on privacy protection model mentioned earlier,the algorithm is designed to publish anonymous datasets,which includes the ExtractRelation algorithm to extract the individual correlation and ProtectionAlgorithm algorithm to generate safe anonymous publishing table.And also give the theoretical analysis of the algorithm's correctness,security,availability and cost.Finally,experiment is designed for the algorithm,and experimental results show that the algorithm in our paper can solve the attack based on individual correlation.Compared with m Invariance protection algorithm and sequential release ? diversity algorithm,the privacy protection algorithm proposed in this paper not only has similar usability,but also has better security.