增量加权社会网络隐私保护技术

发布时间：2019-05-26 23:15

【摘要】：随着互联网与信息化技术的迅速发展,社会网络逐渐普及,越来越多的人们在网上共享数据并参与各种各样的活动。社会网络的出现无疑给人们的沟通与娱乐提供了及其广阔的平台,但随之而来的是大量包含个人隐私信息的社会网络数据被发布到网络上,恶意攻击者可以利用多种背景知识进行隐私攻击,从而导致用户隐私信息的泄露。因此如何保护社会网络中的隐私信息已成为当前数据隐私保护研究领域的热点问题,近年来出现了多种社会网络匿名化技术。实际生活中存在大量增量变化的社会网络,即随着时间的迁移网络中结点和边不断增加,例如email通信网络,另外,在进行社会网络分析时发现大部分社会网络中结点之间的边是带有权重信息的,即存在许多加权社会网络图,加权图与简单图相比携带了更多社会网络中的信息,也会带来更多的隐私泄露。因此,本文重点研究支持增量加权社会网络的隐私保护技术。本文首先综述了现有的社会网络隐私保护技术,并基于此,将社会网络抽象成加权图增量序列,并针对加权图增量序列中结点身份泄露问题和边权重泄露问题进行研究。本文首先定义了增量序列分类安全条件(Increment Sequence Class Safety Conditio n,缩写ISCSC)来指导匿名过程,以解决加权图增量序列中结点身份泄露和边权重泄露问题,并证明了满足ISCSC是实现隐私目标的必要条件。提出了基于权重链表的k-匿名隐私保护模型,并设计了满足基于权重链表的k-匿名模型的WLKA算法,以防止结点身份泄露,从而有效的防止了基于结点标签及权重链表的隐私攻击。提出了基于超图的k-匿名隐私保护模型,并设计了满足基于超图的k-匿名模型的HVKA算法,以进一步保护边权重的安全性并提高发布图数据的可用性,有效的防止了基于结点标签的隐私攻击。最后,在真实数据集上进行了大量的测试研究,通过实验结果本身及对实验结果的分析,证明了WLKA算法能够有效的防止结点身份泄露；HVKA算法则在保证结点身份和边权重信息安全性的同时更好的保留了原图的结构性质并提高了权重信息的可用性,同时还降低了匿名过程的时间代价。
[Abstract]:With the rapid development of Internet and information technology, social network is becoming more and more popular. More and more people share data and participate in a variety of activities on the Internet. The emergence of social network undoubtedly provides people with a broad platform for communication and entertainment, but a large number of social network data containing personal privacy information are released to the network. Malicious attackers can use a variety of background knowledge for privacy attacks, resulting in the disclosure of user privacy information. Therefore, how to protect the privacy information in the social network has become a hot issue in the field of data privacy protection. In recent years, a variety of social network anonymity technologies have emerged. There are a lot of incremental social networks in real life, that is, the number of nodes and edges in the migration network increases with time, such as email communication network, in addition, In the analysis of social network, it is found that the edges between nodes in most social networks have weight information, that is, there are many weighted social network graphs, and the weighted graphs carry more information in social networks than simple graphs. There will also be more privacy disclosures. Therefore, this paper focuses on privacy protection technology that supports incremental weighted social networks. In this paper, the existing privacy protection technologies of social networks are reviewed, and based on this, the social networks are abstracted into weighted graph incremental sequences, and the node identity disclosure problem and edge weight leakage problem in weighted graph incremental sequences are studied. In this paper, the security condition (Increment Sequence Class Safety Conditio n, abbreviation ISCSC for incremental sequence classification is defined to guide the anonymous process in order to solve the problem of node identity leakage and edge weight leakage in weighted graph incremental sequence. It is proved that satisfying ISCSC is a necessary condition to achieve the goal of privacy. A k-anonymous privacy protection model based on weight linked list is proposed, and a WLKA algorithm satisfying the k-anonymity model based on weight linked list is designed to prevent node identity disclosure. Thus, the privacy attack based on node label and weight linked list is effectively prevented. A k-anonymity privacy protection model based on hypergraph is proposed, and a HVKA algorithm based on hypergraph is designed to further protect the security of edge weight and improve the availability of published graph data. It effectively prevents privacy attacks based on node tags. Finally, a lot of test research is carried out on the real dataset. Through the analysis of the experimental results and the experimental results, it is proved that the WLKA algorithm can effectively prevent node identity leakage. HVKA algorithm not only ensures the security of node identity and edge weight information, but also preserves the structural properties of the original graph and improves the availability of weight information, and at the same time reduces the time cost of anonymous process.
【学位授予单位】：东北大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.08

【相似文献】