基于社交网络的隐私保护技术研究与应用

发布时间：2018-04-02 06:34

本文选题：社交网络　切入点：隐私保护　出处：《中国科学院研究生院(沈阳计算技术研究所)》2016年硕士论文

【摘要】：社交网络即SNS,作为Web2.0的技术产物之一,已经成为人们在互联网上传播信息、沟通交流的主要平台。它的主体是用户和用户之间的相互关系,通过各种行为对这种用户关系进行维系和巩固。社交网络是一个信息共享为目的的平台,保存了用户大量的属性信息和好友关系信息。随着社交网络应用的普及,用户个人信息和关系数据的累积,在信息共享的前提下保护数据的隐私是一项很重要的挑战。本文首先研究了社交网络的特征和隐私保护方面面临的威胁。结合目前的社交网络隐私保护技术现状,分别在用户属性信息和用户关系信息两个方面进行了隐私保护设计。针对用户属性的隐私保护,本文通过应用数据挖掘中泛化和隐匿理论,提出SNS细粒度属性匿名算法实现了用户自主选择属性的隐私保护程度的需求。由于减少了匿名化的属性数目,与原始算法相比该算法在SNS网络中的时间复杂度明显降低。针对关系数据隐私性保护和可用性的需求,根据传统的边随机化算法,本文提出了一种安全、高效的社交网络隐私保护及其权限控制方案。本文提出的基于群的节点划分和基于群的边随机化算法,利用图特征显著性约束得出该算法不仅提高了社交网络的隐私保护程度也保证了社交网络结构的可用性。结合该算法提出的权限控制方案,以权限传递的方式加快了权限判定的效率。最后,通过搭建实验仿真平台,测试了细粒度属性匿名算法的运行时间和信息损失量、基于群的边随机化算法的特征显著性和用户访问权限效率。分析结果表明,与现有社交网络隐私保护技术相比,本文所提算法和方案能高效的保护社交网络数据的隐私并能保证社交网络图结构上的可用性,实现了用户属性和关系数据既保护又共享的需求。
[Abstract]:The social network SNS, as one of the products of Web2.0 technology, people on the Internet has become the main platform for the dissemination of information and communication. It is the main body of the relationship between the user and the user, the user relationship to maintain and consolidate through various behavior. Social network is an information sharing platform for the purpose of preservation. The information and attribute information of friends a lot of users. With the popularity of social networking applications, the accumulation of users' personal information and relationship data, in the premise of information sharing under the protection of the privacy of the data is a very important challenge. This paper studies the face features and privacy protection of social network threats. Combined with the social network the current technology status of privacy protection, privacy protection design was carried out in two aspects of user attribute information and user information. According to the relationship between privacy of user attributes This article through the protection, generalization and suppression theory used in data mining, put forward the fine-grained SNS attribute anonymity algorithm to achieve the user to choose the attribute of the degree of privacy protection requirements. Due to the reduction of the number of properties anonymilization, compared with the original algorithm the algorithm in SNS network complexity is significantly reduced. According to the relationship between data privacy protection and availability requirements, according to the random algorithm, this paper proposes a security, social network privacy protection and access control scheme with high efficiency. Based on the nodes into groups and randomized algorithms based on edge group, with significant constraints that features of the algorithm not only improves the privacy protection degree of social the network also guarantees the availability of social network structure. Combined with the proposed access control scheme is adopted to accelerate the transfer to access permissions to The efficiency. Finally, through building simulation platform, operation time and information loss of fine-grained attribute anonymity algorithm to test the edge of randomized algorithms group significantly and user access efficiency. Based on the analysis results show that, with the existing social network privacy protection technology in the social network data privacy protection algorithm and scheme can efficiently and can ensure the availability of the structure of the social network on the map, to achieve the user attribute and relationship data protection and sharing needs.

【学位授予单位】：中国科学院研究生院(沈阳计算技术研究所)
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP309

【相似文献】