Achieving k-anonymity base on nearest neighbor search

作者：Yang, Gaoming^*; Li, Jingzhao; Zhang, Shunxiang; Zhu, Guangli

来源：Journal of Information and Computational Science, 2013, 10(12): 3971-3979.

DOI：10.12733/jics20102082

• Clustering technologies
• Data publishing
• Data utilities
• Information loss
• K-Anonymity
• Nearest Neighbor search
• Privacy preserving
• Similar datum

Using generalization/suppression technology to achieve k-anonymity is a simple and efficient method, but this method can lead to an over generalization problem. Since the object of data release is for the other to uses, over generalization often leads to data useless. To minimize the information loss and increase data utility for anonymous data set, it is crucial to partition similar data together using clustering technology and then to anonymize each cluster individually. For this, we propose a method using nearest neighbor search to achieve k-anonymity, and the anonymous dataset must maintain as much information as possible. Experimental results show that our method outperform previous proposed clustering algorithms in information loss and running time.

• 出版日期2013
• 单位安徽大学