Researeh On Sensitive Information Preserving In Data Sharing

Along with the rapid progress of computational capability, storage technology and network technology, national informationalization construction has gotten great development. Information comes from dada, so that data sharing is a key step during informationalization construction. For example, real-name in civil aviation and train ticket system could be implemented smoothly benefits from their sharing data with household registration system. Data sharing has benefit our country or enterprises or even individual a lot so far, but privacy disclosure problems following are so serious that could not be ignored. People would not like to share their data for being afraid of revealing their privacy, which becomes an obstacle during the construction of national informationalization. Therefore, how to protect private data while sharing data has important theoretical and realistic significance. Related research has also became a hot subject in the filed of database and data mining.This paper expounded private preserving technology at first. Then we considered how to protect private data while sharing original data sets. After that, we studied how to hind sensitive patterns in frequent pattern sharing. The main contributions are as follows:(1) Studied deeply on privacy preserving in data sharing, especially on how to preserve privacy by data sanitization. Summarized and described the principle and process of data sanitization.(2) Studied on how to preserve privacy while sharing original data sets, and put forwards a data sanitization algorithm based on weak sensitive pattern tree. Then designed experiments to test the sanitization effects of the proposed algorithm Dpur and existing algorithm SWA, and analyzed the experimental results at last.(3) How to preserve privacy while sharing frequent patterns is studied at last. Considering preserving privacy by hindering existence of sensitive patterns, proposed a pattern sanitization algorithm based on sensitive item. Then designed experiments to test the effectiveness and efficiency of those algorithms,and analyzed the experimental results.