Abstract
Data mining is the progress of automatically discovering high level data and trends in large amounts of data that would otherwise remain hidden. In order to improve the privacy preservation of association rule mining, a hybrid partial hiding algorithm (HPH) is proposed. The original data set can be interfered and transformed by different random parameters. Then, the algorithm of generating frequent items based on HPH is presented. Finally, it can be proved that the privacy of HPH algorithm is better than that of the original algorithm.
Secure Sum based Privacy Preservation Association Rule Mining on Horizontally Partitioned Data