Anonymous privacy protection method for multi-sensitive attribute data release

Abstract

The invention discloses an anonymous privacy protection method for multi-sensitive attribute data release. The method comprises the steps of 1, calculating correlation among attributes in an originaldata table, and defining attribute categories; 2, grouping all the attributes according to the correlation among the attributes; 3, sorting the groups from high to low according to the correlation among the attributes; 4, calculating the occurrence frequency of the sensitive attribute values in each group, and judging whether the attribute values need to be protected or not; 5, exchanging the sensitive attribute value needing to be protected and the associated attribute thereof with the sensitive attribute value needing not to be protected; 6, performing frequent item set mining on the relatedattributes; and step 7, generating an anonymous equivalent group by using a clustering technology, wherein each sensitive attribute in the anonymous equivalent group satisfies L diversity. By reducing the anonymity rate and jointly replacing the associated attributes, the availability of the data is improved while the privacy protection strength is ensured.

Description

technical field [0001] The invention belongs to the technical field of multi-sensitive attribute data; in particular, it relates to an anonymous privacy protection method for publishing multi-sensitive attribute data. Background technique [0002] In recent years, with the rapid development of the Internet, the dependence on information sharing in various fields has generally increased. Data collectors can often find more valuable potential information by mining shared data. However, there is a privacy leak in the process of information sharing. Risk, since the original data contains a lot of sensitive information, such as home address, medical information, work status, etc., if the data owner directly releases the original data, it will lead to the leakage of personal privacy. Therefore, privacy protection for the original data is required Post after processing. [0003] With the rapid increase of the amount of data, the structure of the data table has become more complica...

AI Technical Summary

Problems solved by technology

[0006] For multi-sensitive attribute data tables with correlation, the existing privacy protection model cannot take into account the correlation between attributes and the availability of data while realizing the protection of multi-sensitive attribute data, and the final published data table cannot resist all attacks model, existing methods cannot better balance the availability of data and the robustness of privacy protection

Images