Micro-aggregation anonymization method based on sorting

Abstract

The invention provides a micro-aggregation anonymization method based on sorting. The method comprises the following steps that: (1) a sorting operation: on the basis of a Q1 quasi-identifier, dividing a dataset into a plurality of categories to enable k-division to be based on the Q1 quasi-identifier by the dataset; (2) a division operation based on sorting: independently systemically forming equivalence classes from the first extreme record and the last extreme record of the dataset initialization of the sorting operation, and keeping the record number of the equivalence classes in k; and (3) an aggregate operation: taking the center points of two extreme records as the centroid point of each equivalence class, and replacing all sensitive attribute values with the mean value of the equivalence classes to form an anonymous equivalence class. By use of the method, firstly, according to a mean valve sorting technology, a k-division process is effectively improved to guarantee the information loss ratio of the k-division process to be minimum, the execution efficiency of an algorithm is improved, in addition, a multidimensional dataset can be processed after a sorting concept is introduced, and then, privacy protection can be improved.

Description

technical field [0001] The invention relates to a sorting-based micro-aggregation anonymization method, belonging to a data privacy protection method in the field of information security. Background technique [0002] In the current information age, the release of data is beneficial to the field of data analysis, such as the release of election ballot information, census information, medical and health information, etc. Through the mining of such information, the trend of election information and the trend of population growth can be judged and the health of the people. But these released data more or less contain some sensitive information, such as personal privacy. To avoid the threat of data privacy breaches, data should be anonymized before release. [0003] In the current anonymization field, the most widely used anonymity technology is k-anonymity, whose main idea is to make each record indistinguishable from at least other k-1 records. The general implementation pr...

AI Technical Summary

Problems solved by technology

[0006] Purpose of the invention: Aiming at the inability to realize the multidimensional data set processing and the slightly high information loss rate in the prior art, the present invention combine...

Images