Research on k-anonymity technolgy based on R-tree
Chu Yan-hua · Journal of Inner Mongolia University of Science and Technology · 2010
Privacy protection algorithms are currently used for proceeding anonymous static data sets,while most of the frequently contacted data sets are dynamic.Therefore,research of privacy protection algorithm for dynamic date sets is a hot issue.k-anonymity technology based on R-tree is suitable for dynamical changing data sets.During the process of building R-tree,original two-splitting algorithm can not guarantee the similarity of the child node.In this paper,k-means multi-splitting algorithm was proposed to replace the original two-splitting algorithm,and the problem of similarity of child node with increasing the value of k under fixed M was solved.The results of experiment showed that by merging the k-means multi-split algorithm,R-tree has effectively improved the k-anonymous quality and the query efficiency of the anonymous forms.