基于DSNPP算法的社交网络隐私保护方法

Privacy-preserving Approach in Social Networks Based on DSNPP Algorithm

社交网络发展迅速,数据发布过程中存在的一个重要安全隐患就是隐私泄露。针对目前大多数社交网络隐私保护研究存在的"人员属性隐私保护"和"社区结构保护"之间没有实现真正结合的问题,就两者综合考虑,提出一种基于密度聚类算法的社交网络隐私保护方法(Density for Social Network Privacy-Preserving,DSNPP)。该算法通过对节点进行密度聚类分析,得到任意形状的簇,采用对簇内节点进行泛化、在簇内插入真实节点、增加相应边等技术来保护节点的信息和节点之间的关系信息,从而实现了人员属性隐私保护和社区结构保护两方面的真正结合。最后,通过实验表明,与p-Sensitive k-匿名算法、GSNPP算法相比,该算法信息丢失量上优势明显,可以获得更高的隐私保护。

With the rapid development of social network, an important safety hazard that exists in the process of data publishing is leakage. For the questions that most researches on social network privacy protection do not realize existence of protecting privacy in propexty and community structures,considering the both, propose a method of social networking privacy, Density for Social Network Privacy-Preserving （DSNPP）. The algorithm is based on density clustering method, which gets clusters in arbitrary shape through nodes cluster analysis, and it uses the technology of generalizing cluster nodes,inserting the real nodes in the cluster,increasing corresponding edges and so on to protect information of nodes and the relationship between nodes, which achieves purpose of social networks privacy protection. Finally,compared with p-Sensitive k -anonymous algorithm and GSNPP algorithm, the algorithm has the advantage in the amount of information loss, and it can obtain higher privacy protection.