Multi-sensitive attribute privacy protection method based on intra-group rearrangement

Abstract

The invention belongs to the technical field of information security, and discloses a multi-sensitive attribute privacy protection method based on intra-group rearrangement. The method comprises steps of firstly, measuring correlation between SA and SA attribute values by calculating the lifting degree of various value combinations between every two columns of sensitive attributes SA, if the lifting degree is larger than 1, indicating positive correlation, and enabling the value combinations with the correlation to be hidden until the lifting degree is reduced to 1 or below; then, calculating chi-square values between the QI and the SA attributes by counting values between each column of quasi-identifier attributes QI and each column of SA attributes, checking a chi-square distribution table to judge whether correlation exists or not, and stopping checking the QI when at least one column of SA is related to the current QI, so that the QI attributes are divided into two types which are related to the SA and are not related to the SA; and then performing multi-dimensional bucket grouping on the data table to form a group which enables each SA to meet L diversity, and finally performing intra-group rearrangement on the QI columns having correlation with SA attributes.

Description

technical field [0001] The invention relates to a multi-sensitive attribute privacy protection method based on group rearrangement, which belongs to the technical field of information security. Background technique [0002] Privacy Preserving Data Publishing (PPDP) is a promising approach to information sharing that can simultaneously protect individual privacy and sensitive information. Fields such as finance, medical care, and e-commerce are inseparable from the release of a large amount of data. In daily life, the use of multi-dimensional sensitive data (MSA) is more common than single-dimensional sensitive data (SSA). However, the shared use of data also brings various privacy leakage problems, which adds many security risks to people's daily life. At present, most privacy protection methods are privacy models formulated for the leakage of single sensitive attributes, such as the three basic models of k-anonymity, l-diversity, and t-approximation, which are mainly impl...

AI Technical Summary

Problems solved by technology

[0005] In order to effectively solve the problem of leakage of correlation information between attributes of multi-sensitive attribute data, the present invention proposes a privacy protection method based on intra-group rearrangement of multi-sensitive attributes

Images