中文摘要：

针对含敏感关系的社会网络中用户隐私信息泄露的问题，结合攻击者基于背景知识的多种类型攻击，提出一种（k^2，l）-匿名模型，并分别通过基于动态规划和贪心算法的度序列匿名算法实现（k^2，l）-匿名模型。该模型可同时抵御社会网络中敏感关系识别攻击、节点度攻击和朋友连接攻击。在数据集上进行仿真实验，通过分析对比实现（k^2，l）-匿名模型的两种算法，分析对比结果表明了（k^2，l）-匿名模型具有较高的匿名质量和较低的信息损失，能有效地保护含敏感关系社会网络中用户的隐私信息。

英文摘要：

For user’s privacy information disclosure in social networks with sensitive relationship,combined with various types of attacks based on background knowledge of attackers,the （k^2 ,l）-anonymous model was presented,and the （k^2 ,l）-anonymous model was implemented using degree sequence of anonymous algorithm based on dynamic programming and greedy algorithm respectively.Finally,theoretical analysis and contrasting results of the two algorithms on the data sets verify the higher quality and lower anonymous information loss of the model and it can protect user’s privacy in social networks with sensitive relationship.