K-anonymous privacy protection method based on quasi-identification attribute value frequency

Abstract

The invention provides a k-anonymity privacy protection method based on quasi-identity attribute value frequency which is beneficial to data analysis and information loss reduction. The anonymous privacy protection method is characterized by comprising the following steps of: 1) for each quasi-identification attribute in a to-be-published data table, establishing a generalized hierarchical tree ofeach quasi-identification attribute according to a data distribution condition and a service condition; 2) calculating the attribute value number, the attribute value frequency and the attribute value entropy of each quasi-identification attribute; 3) selecting a quasi-identification attribute to carry out hierarchical global generalization according to a priority sequence of minimum attribute value frequency, maximum attribute value number and minimum attribute value entropy; And 4) carrying out k-anonymous model inspection on the to-be-published data table subjected to global generalizationin the step 3), if the k-anonymous model is met, taking the to-be-published data table subjected to global generalization in the step 3) as a final published data table, otherwise, repeating the step2), the step 3) and the step 4).

Description

technical field [0001] The present invention relates to the field of data privacy protection, and more specifically, to a k-anonymous privacy protection method based on the value frequency of quasi-identification attributes. Background technique [0002] With the development of social informatization and networking, data has shown explosive growth, and the era of big data has arrived; at present, big data has become a research hotspot in academia and industry, and is affecting people's daily life, work habits and ways of thinking ;In data publishing applications, how to protect the privacy of data users has become a major challenge; the easiest way to protect user privacy is to directly delete the identification attributes such as names and ID numbers when publishing original data to prevent users from Privacy leakage; however, because some attribute value sets in the dataset released in this way can accurately or more accurately re-identify an individual with relevant infor...

AI Technical Summary

Problems solved by technology

However, the information loss of global generalization is generally greater than that of local generalization; therefore, how to make the published data table satisfy the k-anonymity model while facilitating data analysis and reducing information loss has become an important issue in privacy protection research. focus

Images