Research On Privacy-Preserving Data Mining

During the net information times , information technology makes the process of data collection and transmission become more and more user-friendly ,showing an explosive data growth. The degree of concern about privacy is becoming increasingly high. Data mining is technology of useful knowledge and rules from the mass of data. Therefore, the abuse of data mining technology is increasingly likely,consequently ,how to cross the gap between data mining requirements and data security and provide effective privacy protection would be the focus to solve the problem of privacy preserving data mining .This paper analyzes the lack of definition of the privacy object protection in the existing methods , and further illustrates the fact that relationship between the data constitutes the object of privacy protection. Protection of one of them separatly can not be disclosed to avoid the possibility of privacy. In the new model of privacy protection based on the object, a targeted strategy for privacy protection is presented : decomposition of the relationship between the data could truly protect the privacy of the data. In this study, analysis of existing data and the traditional anonymity of the lack of data exchange strategy were proposed based on the uncertainty of the privacy protection technology mapping. This method protects the privacy of the data through increasing data users' after-disturbance uncertainty.The availability of after-disturbance data pertaining to data analysis and data mining meets the needs of users.