Abstract: With the development of the Internet, social network has becoming an important platform of?people chat, make friends, speech. There are a lot of privacy information in the social network, as a result the information which involve the personal identity, social structure, attribute contact and so on constitute the privacy information. Therefore, effective protection should be carried out on social network. Analysis finds that the existing part of the social network privacy protection scheme based on structure disturbance has some drawbacks of high?network structure disturbance and high information losing. This article puts forward a novel privacy protection scheme based on node reconfiguration focusing on solving the problems of the similarity analysis and selection of the edge nodes and protected node, the edge nodes isomorphic?transformation, as well as the optimal disturbance in the original privacy node structure. The test results show that the method can effectively protects the privacy of social networks, also can guarantee the availability of the network structure at the same time.

Key words: social network, identity privacy, edge node, figure structure modification

摘要： 互联网的发展使社交网络成为人们聊天、交友、发表言论的重要平台。在社交网络中会存在大量信息，其中涉及到个人身份，社交结构，属性联系等信息就构成了隐私信息，因此需对隐私信息进行有效的保护。分析发现，现有部分基于结构扰动的社交网络隐私保护方案存在对网络结构扰动大、信息丢失率高等弊端。为此，提出了一种基于节点重构的隐私保护方案，重点解决了边缘节点与受保护节点的相似性分析与选取，边缘节点的同构化改造，以及最优化扰动原隐私节点结构等问题。实验结果表明提出的方法在有效保护社交网络隐私的同时，还能确保信息的高可用性。

关键词: 社交网络, 身份隐私, 边缘节点, 图结构修改