Privacy protection-based node triangle counting publishing method for large graph

Abstract

The invention discloses a privacy protection-based node triangle counting publishing method for a large graph, which comprises the following steps of: selecting an original graph G, and counting triangle parameters of each node in the original graph G by using a triangle counting algorithm to obtain a first distribution histogram; counting triangular parameters of each node in the preprocessed graph G theta by using a triangular counting algorithm to obtain a second distribution histogram; and adding a Laplace noise protection node to the data bucket to solve the privacy leakage problem of triangle counting related data in the publishing process. According to the invention, unnecessary edges are deleted through certain preprocessing means, the sensitivity upper limit of the graph is controlled within a certain range, and a better histogram publishing method is selected, so that the noise amount required to be added for meeting differential privacy can be greatly reduced, and the optimal balance between the privacy and the availability of published data is achieved.

Description

technical field [0001] The invention relates to the technical field of differential privacy, in particular to a publishing method of node triangle counts of large graphs based on privacy protection. Background technique [0002] The concept of "differential privacy" was first proposed by Dwork et al. in 2006. Its definition can be simply described as: if there are two adjacent data sets with a difference of at most one record, the attacker simultaneously Query, the query results obtained can not track this record. Dwork also proposed a specific method to achieve differential privacy: adding noise that satisfies the Laplacian distribution can achieve differential privacy. Since then, some scholars have proposed an exponential mechanism. Compared with the Laplace mechanism, it can provide a larger number of queries under the same privacy budget setting, but at the same time it will bring greater computational complexity and more long computation time. [0003] "Number of tr...

AI Technical Summary

Problems solved by technology

The existing combined query of triangular calculation and node count still directly deletes edges and directly adds noise to the graph. Therefore, the accumulated noise is too large and the usability of the query results is poor, which is not an ideal solution.

Images