Differential privacy data publishing method and system based on dependency removal

Abstract

The invention discloses a differential privacy data publishing method and system based on dependency removal. Each data owner independently sections an own original data set and obtains a differential value data set, the first two items of the differential value data set are subjected to summation and noise adding, then, the obtained value is uploaded to a cloud service provider after the obtained value is subjected to homomorphic encryption, and the cloud service provider carries out homomorphic decryption operation; each data owner transforms the differential value data set item to obtain a transformation coefficient, and stable noise consisting of white noise which is mutually independent and complies with Gaussian distribution is added; the data owner carries out inverse wavelet transform to obtain a scrambled data set, and the data set is uploaded to the cloud service provider; and the cloud service provider utilizes Wiener filtering to carry out filtering, and the cloud service provider carries out inverse transformation on a refined data set and publishes the refined data set to the data item of a third-party data user when the data user requests the data set. The differential privacy data publishing method and system effectively reduces the data calculation amount and interaction amount and improves a resource use ratio and data availability.

Description

technical field [0001] The present invention relates to the field of data privacy protection under the cloud computing environment, and more specifically, relates to a method for publishing homomorphic differentially private data based on correlation removal, which is used to solve the large amount of calculation and complicated interaction of similar differentially private protection methods in the process of data publishing The problem. Background technique [0002] With the rapid development of cloud computing and mobile Internet technology and applications, data sharing applications have penetrated into various fields. While data sharing applications have brought about a huge improvement in service quality, the accompanying data privacy security issues have become increasingly severe. While not completely trustworthy cloud service providers bring efficient data services to users, they also have the possibility of coveting users' private data and causing user privacy leak...

AI Technical Summary

Problems solved by technology

[0009] The purpose of the present invention is to propose a high-efficiency differential privacy data publishing method based on correlation removal to reduce computational complexity and interaction volume in order to solve the defects of large amount of calculation and frequent interaction in the existing similar differential privacy correlation removal technology in the cloud environment , thus increasing the efficiency

Images