Privacy Preserving Based On Dynamic Datasets

With the rapid development of information technology, information on the Internet become more and more chaotic, people on the Internet to obtain needed information has become more and more difficult, because the Internet provides useful information to people while also providing people with a lot of redundant information. So, now, to obtain information from the mass of information to their advantage increasingly important to people's work, this, experts and scholars to open up the data mining research, the purpose is to study how to efficiently complete computer search for information in order to get people useful information. Currently, data mining has demonstrated its value, the emergence of various search engines, and all kinds of information mining technology to solve many problems for people. However, as data mining applications, there is status quo is of growing importance, namely, loss of privacy issues. So far, literature data sets for static and dynamic data sets to study privacy issues. This paper pointed out the lack of existing literature on the basis of the algorithm, to solve the problem of the yet to propose solutions. The main research works are as follows:1) For the more privacy sensitive data sets proposed protection algorithm. Although the existing literature for two types of static and dynamic data sets, but they it is assumed that these two data sets have only one sensitive attribute, and we in real life, a data set often have multiple sensitive attributes, analyzing multi-attribute data sets sensitive characteristics, the design of a multi-dimensional array grouping, while the technology based on the proposed three greedy grouping strategy more sensitive to the privacy of handling attribute data sets. Finally, the experiment by changing the three parameters, namely, the data set size,1 parameter and the number of sensitive property, observe the efficiency of three algorithms.2) Multi-attribute data sets sensitive solutions to the problems re-release. Existing literature on the Dynamic Data Sets released studied but are based on single-sensitive data sets, this article draws on their algorithms to the idea of them improved and more sensitive data sets used in the data re-release problem, has been discussed earlier combined with multi-attribute data sets sensitive to the privacy protection algorithm, we propose a new and more sensitive data sets re-released issue of privacy protection algorithm.This paper proposed a solution, it is also the risk of disclosure of privacy given the estimation method proposed SCG map concept behind the algorithm for the understanding of the tools. Finally, experiments for each algorithm, verify the proposed algorithm is feasible.