Clustering-based Data Publishing for Differential Data Anonymization
Liu Ha · Journal of Hainan Normal University · 2014
The demand for various types of data protection are different in a large number of data release process involving the privacy of individuals and institutions. Most of them don't need to be protected, only part of them need to be protected, and the protection needs and different. This method first clustered the tuples which don't need to be protected, and then clustered the tuples which need to be protected according to their different requirements and distance from the cluster. Finally, through the experiment we proved that the method can achieve better classification privacy protection effection.