社交网络中用户隐私信息优化保护仿真研究

发布时间：2018-08-03 15:27

【摘要】：对社交网络中用户隐私信息优化保护,可保证网络用户的信息不被泄露。进行隐私信息保护时,应对隐私信息节点聚类,统计丢失数据信息,并对聚类后的社交网络信息进行匿名化和归一化处理,但是传统方法通过建立形式化验证和基于信任的隐私信息流模型,融合隐私信息访问粒度控制完成对社隐私保护,但是不能对隐私信息节点聚类,无法准确统计丢失的数据信息和进行匿名处理,导致出现隐私信息保护效果不好的问题,提出一种基于k-邻域同构的社交网络中用户隐私信息优化保护方法。上述方法融合于三角矩阵理论映射出社交网络中的节点邻域子图,给出相邻隐私信息时间片间的关联性,依据信息变化增量确定基准节点,在此基础上对社交网络中隐私信息节点进行聚类,对隐私信息节点生成的簇进行簇内外泛化,对社交网络进行匿名化处理,统计出丢失的各类型信息,进行归一化,完成对社交网络中用户隐私信息的优化保护。仿真结果证明,所提方法有效地防止隐私泄露,增强了数据发布的安全性。
[Abstract]:The optimal protection of user privacy information in social networks ensures that the information of the network users is not leaked. When the privacy information is protected, the privacy information nodes should be clustered, and the data are lost, and the social network information after the clustering is anonymously and normalized, but the traditional method establishes a formal verification and basis by establishing a formal verification and basis. The trusted privacy information flow model, which integrates privacy information access granularity control to complete the protection of social privacy, can not cluster the privacy information nodes, can not accurately statistics the lost data information and carry out anonymous processing, which leads to the problem of the poor protection effect of privacy information, and proposes a social network based on k- neighborhood isomorphism. This method combines the triangular matrix theory to map the neighborhood sub map of the social network, gives the correlation between the time slices of the adjacent privacy information, and determines the datum node according to the increment of the information change. On this basis, the privacy information node in the social network is clustered to the privacy information node. The generated clusters are generalised inside and outside the cluster, and the social network is anonymously processed, the lost types of information are counted, and the user privacy information in the social network is optimized. The simulation results prove that the proposed method can effectively prevent privacy disclosure and enhance the security of data publishing.
【作者单位】：南阳理工学院软件学院;
【基金】：河南省科技厅基础前沿项目(142300410108)
【分类号】：TP309;TP393.09

【相似文献】