脆弱好友检测模型的研究与实现
发布时间：2018-11-06 20:31
【摘要】：随着Web 2.0时代的到来,用户被在线社交网络(Online Social Networks,OSNs)呈现出的“低门槛、高开放性”等特点所吸引,参与意识也被无限激发。作为信息生产者,用户为社交网络提供了大量的信息,包括用户个人信息(Personal Preference Profiles,PPPs)和用户原创内容(User Generated Contents,UGCs)。在社交网络的大数据环境下,这些碎片化的信息将被整合利用。然而,用户个人信息和用户原创内容往往会涉及到用户的隐私,而大数据的挖掘能力恰恰威胁到了用户的隐私保护。从根本上来说,隐私泄露源自于用户以及他所处社交圈的脆弱性,具体体现为用户自身隐私保护意识的薄弱和社交圈中用户好友的隐私传播行为及其恶劣影响。尤其是作为信息传播过程中强有力的推动者,用户好友没有充分意识到自己在隐私信息泄露中所扮演的威胁性角色。因此,研究用户好友在隐私信息传播过程中的影响作用,对于从根本上保护用户隐私有着至关重要的意义。本论文在总结现有的研究成果的基础上,首次提出了传播脆弱性的概念,并综合考虑了用户好友的传播脆弱性在各个方面对隐私保护的影响,结合动态隐私信息传播的特性,构建了隐私接收扩散模型(Privacy Receiving-Disseminating Model,PRD)并实现了相应的最终扩散圈(Ultimate Circle of Disseminating,UCD)迭代算法,从崭新的视角,即动态隐私信息的传播过程,去评估用户好友的传播脆弱性,最终研究并实现了一个新颖的脆弱好友检测模型(Vulnerable Friend Identification Model,VFI)。VFI 模型能够帮助用户从众多直接好友中识别出脆弱好友,认清他所处社交圈的脆弱性,并通过解除好友关系等措施保护用户的个人隐私安全。论文首先介绍了社交网络中隐私保护问题的研究现状和相关技术;之后整体阐述了基于动态隐私信息传播过程的脆弱好友检测方案的研究与设计;接下来详细描述了方案中各个模块的具体设计与实现,包括传播脆弱性量化模块、隐私接收扩散模块、以及脆弱好友检测模块;并利用Facebook和Twitter的真实数据验证了 VFI模型的有效性及其优良性能;最后总结全文,对未来工作进行了展望,并总结了作者在研究生期间的工作和成果。
[Abstract]:With the arrival of the Web 2.0 era, users are attracted by the features of "low threshold, high openness" presented by (Online Social Networks,OSNs (online social network), and the consciousness of participation is aroused indefinitely. As an information producer, users provide a large amount of information for social networks, including user personal information (Personal Preference Profiles,PPPs) and user-generated content (User Generated Contents,UGCs). In the social network big data environment, these pieces of information will be integrated into the use of. However, user's personal information and user-generated content often involve user's privacy, and big data's mining ability just threatens the user's privacy protection. Especially as a powerful promoter in the process of information dissemination, users' friends are not fully aware of their threatening role in the disclosure of privacy information. Therefore, it is very important to study the influence of user friends in the process of privacy information dissemination. On the basis of summarizing the existing research results, this paper puts forward the concept of communication vulnerability for the first time, and synthetically considers the impact of the communication vulnerability of user friends on privacy protection in all aspects, combining with the characteristics of dynamic privacy information dissemination. The privacy receiving diffusion model (Privacy Receiving-Disseminating Model,PRD) is constructed and the corresponding iterative algorithm of the final diffusion circle (Ultimate Circle of Disseminating,UCD is implemented. From a new perspective, the propagation process of dynamic privacy information is introduced. Finally, a novel fragile friend detection model, (Vulnerable Friend Identification Model,VFI). VFI model, is developed to help users identify vulnerable friends from a large number of direct friends. Recognize the vulnerability of his social circle and protect the privacy of the user through such measures as breaking off his friends. Firstly, this paper introduces the research status and related technologies of privacy protection in social networks, and then describes the research and design of fragile friend detection scheme based on dynamic privacy information dissemination process. Then, the detailed design and implementation of each module in the scheme are described in detail, including the dissemination vulnerability quantification module, the privacy receiving diffusion module and the fragile friend detection module. The validity of the VFI model and its excellent performance are verified by using the real data of Facebook and Twitter. Finally, the paper summarizes the full text, looks forward to the future work, and summarizes the author's work and achievements during the post-graduate period.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP393.09;TP309
本文编号：2315388
[Abstract]:With the arrival of the Web 2.0 era, users are attracted by the features of "low threshold, high openness" presented by (Online Social Networks,OSNs (online social network), and the consciousness of participation is aroused indefinitely. As an information producer, users provide a large amount of information for social networks, including user personal information (Personal Preference Profiles,PPPs) and user-generated content (User Generated Contents,UGCs). In the social network big data environment, these pieces of information will be integrated into the use of. However, user's personal information and user-generated content often involve user's privacy, and big data's mining ability just threatens the user's privacy protection. Especially as a powerful promoter in the process of information dissemination, users' friends are not fully aware of their threatening role in the disclosure of privacy information. Therefore, it is very important to study the influence of user friends in the process of privacy information dissemination. On the basis of summarizing the existing research results, this paper puts forward the concept of communication vulnerability for the first time, and synthetically considers the impact of the communication vulnerability of user friends on privacy protection in all aspects, combining with the characteristics of dynamic privacy information dissemination. The privacy receiving diffusion model (Privacy Receiving-Disseminating Model,PRD) is constructed and the corresponding iterative algorithm of the final diffusion circle (Ultimate Circle of Disseminating,UCD is implemented. From a new perspective, the propagation process of dynamic privacy information is introduced. Finally, a novel fragile friend detection model, (Vulnerable Friend Identification Model,VFI). VFI model, is developed to help users identify vulnerable friends from a large number of direct friends. Recognize the vulnerability of his social circle and protect the privacy of the user through such measures as breaking off his friends. Firstly, this paper introduces the research status and related technologies of privacy protection in social networks, and then describes the research and design of fragile friend detection scheme based on dynamic privacy information dissemination process. Then, the detailed design and implementation of each module in the scheme are described in detail, including the dissemination vulnerability quantification module, the privacy receiving diffusion module and the fragile friend detection module. The validity of the VFI model and its excellent performance are verified by using the real data of Facebook and Twitter. Finally, the paper summarizes the full text, looks forward to the future work, and summarizes the author's work and achievements during the post-graduate period.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP393.09;TP309
【参考文献】
相关硕士学位论文 前1条
1 黎雷;社会网络影响力模型及其算法研究[D];北京交通大学;2010年
，本文编号：2315388
本文链接：https://www.wllwen.com/kejilunwen/ruanjiangongchenglunwen/2315388.html
