Research On K-anonymity Technology Based On R-tree

In today's information society, electronic information and network technology for large-scale information sharing provides a convenient and efficient means. At the same time, privacy protection for new challenges presented. Privacy protection for information-sharing technology research, on the one hand to prevent leakage of sensitive information private to provide strong technical support, and remove information owners to share information in the concern, and promote information exchange and sharing; On the other hand also stressed that the implementation of privacy protection to reduce the losses caused by non-sensitive information and ensure the quality of information sharing, improve the sharing of information availability.Existing privacy protection algorithms for static data set anonymous treatment, but frequent contact with most of the data set is dynamic,the dynamic data set for privacy protection algorithm will be the next hot spot.With the index database technologies become more sophisticated, R-tree spatial index based on the K-anonymity technique, has good scalability and can support the incremental data release, However,for fixed M(R tree node the maximum value of index entries),as the K value increasing, the existing two-splitting algorithm is not related to the specific MBR increment,children below the root node the similarity between the index entries poor, level of privacy protection reduced, affecting the quality of anonymity.Based on k-means algorithm for multi-split R-tree K-anonymity technique, the better to solve that for fixed M, With the continuous increase of K value., the node similar problem, has effectively improved the anonymous quality; Based on k-means algorithm for multi-split R-tree is more reasonable, to reduce the size of the intersection of the MBR, thus improving the query efficiency anonymous table. The experiment proved that, based on k-means multiple R-tree splitting algorithm K-anonymity technique, effectively improve the quality of anonymity and anonymous table of query efficiency.