Social network privacy protection method, system and device

Abstract

The invention discloses a social network privacy protection method. The method comprises the following steps: receiving an input original social network graph; Performing community division on each node in the original social network graph by utilizing a preset label propagation algorithm to obtain each community sub-graph; And reconstructing each community sub-graph by using a sub-graph reconstruction algorithm to complete K-degree anonymization of each community sub-graph so as to realize privacy protection of the original social network graph. According to the technical scheme provided by the invention, the stability of the algorithm is enhanced; the community division effect is improved; the influence of tag selection randomness on tag propagation efficiency and results is reduced; andon the premise of protecting the stability of the social network grid structure, the graph structure is modified to the minimum extent to realize k-degree anonymity, so that the published social network data has relatively good availability. Meanwhile, the invention also provides a system and equipment for protecting the social network privacy and a computer readable storage medium, which have the above beneficial effects.

Description

Technical field [0001] This application relates to the field of privacy protection, and in particular to a method, system, device, and computer-readable storage medium for privacy protection in social networks. Background technique [0002] Social networks belong to the research category of complex networks, focusing on social individuals and their interactions and connections, and are usually modeled as graph data to achieve privacy protection. Attackers often use the background knowledge about the target individual (such as the degree of nodes, identification attributes, connection relationships between nodes, neighborhoods and embedded subgraphs, etc.) to infer individual privacy information. The two types of important private information (node ​​attribute data and connection relationship data) contained in social networks are vulnerable to structural attacks such as node degree attacks and link attacks. [0003] The privacy protection of relational data is a research hotspot t...

AI Technical Summary

Problems solved by technology

However, due to the serious information loss problem in the cluster-based anonymous model after generalization, the network structure changes dramatically and the data utility decreases sharply.

However, most of the anonymization methods for graph data modification or transformation use disturbance methods such as adding, deleting nodes or edges, and subgraph isomorphism to achieve k-degree anonymity. Unable to overcome large information loss problem

Images