Research On Algorithm For Privacy Preserving Based On Association Rule Mining

Recent research has shown that some discovered rules from data mining can pose a threat to security and privacy. As data mining is applied widely in marketing, business, privacy preserving is the research focus in the EC world.Based on the review of privacy preserving, the thesis researches typical algorithms for privacy preserving, builds a new model, propose some improved strategies and a new algorithm. The major research works are listed as follows:(1) Based on the improved process of privacy preserving, this thesis builds a new model of privacy preserving and analyses the former algorithms.(2) Based on the Item grouping algorithm, the improved strategies and a new algorithm is put forward. The major improvements include:Firstly, an inverted table is built by Hash tree. Secondly, a new threshold is introduced.Thirdly, an improved method to select item is put forward which would be removed after clustering the restrictive patterns.(3) From the running time, it is proved theoretically that improved strategies can improve the efficiency of the algorithm based on a comparative analysis between two algorithms.