An Improved K-Anonymity Algorithm Model

Privacy disclosure is a common problem in data publishing, formerly, K-anonymity methods of Privacy Protection have great influence on the data precision. This paper analyzes the reasons of the influence, and proposes an improved algorithm. The algorithm defines a Weight-related of attribute in order to select attributes for generalization. This approach effectively prevents sensitive data loss in the generalization. Experimental results show that the improved algorithm of K-anonymity model increases the data precision effectively.