关键词: 分布式, 社会网络, 隐私保护, 子图匹配

Abstract: Cloud platform is becoming a mainstream tool for storing and processing large-scale social network data. Aiming at large-scale social network subgraph matching privacy protection problem, this paper proposes a privacy preserving algorithm of distributed K-automorphism social network. The algorithm can protect structural privacy of the social network that has been uploaded to cloud platform. The algorithm adds noise edge by passing marked information between vertices, making original graph be anonymous as a K-automorphism social network graph with k symmetric subgraphs. This paper also proposes a distributed subgraph matching method for upload graph. According to the selectivity of nodes in query graph, this paper decomposes the query graph and gets the query decomposition subgraph; distributed and parallel subgraph matching is performed within each computing node to get the matching result of query decomposed subgraph. After the results are joint, the subgraph matching result of query graph is got; in the client, the results of the subgraph matching are recovered and filtered according to the symmetry of the K-automorphism social network graph and the K-automorphism function to get correct matching result. The experimental results show that: distributed K-automorphism social network privacy protection algorithm and distributed subgraph matching method are highly efficient in dealing with large-scale social network graph and these methods can effectively solve the problem of privacy leakage.

Key words: distributed, social network, privacy preserving, subgraph matching