基于语义网的社交网络恶意行为分析

发布时间：2018-05-15 06:46

本文选题：计算机取证 + 社交网络　；参考：《南京邮电大学》2014年硕士论文

【摘要】：社交网络已成为人们日常使用的应用之一，为了提供更好的用户体验，社交网络应用中，可发布各类电子内容，如普通文本、照片、视频等，其中涉及到很多个人隐私方面的信息，这对社交网络应用的安全防护与隐私保护提出了新的挑战。论文首先分析了国内外在社交网络应用安全防护及隐私保护相关问题上研究现状，在相关理论基础的支持下，提出基于本体的社交网络用户关系共享计算模型，以此扩展一对一的用户关系，同时，引入用户好友群关系共享策略，使得黑名单过滤模型得到最大程度的利用。具体而言，论文首先提出一种基于语义网的社交网络黑名单自动过滤本体模型，对现有社交网络本体模型进行定制，并给出一种社交网络恶意信息本体模型的构建方案。然后，利用语义网技术，给出一套解决社交网络黑名单自动过滤的方案。根据现实中信任产生原理，将相关影响因素抽象，给出合理分类标准；引入Pearson算法，以此对用户之间内部以及外部信任产生因素进行量化。在有了黑名单自动过滤模型的基础上，提出一种基于用户群关系的黑名单共享模型，对黑名单过滤模型进行扩展。最后，对上述相关理论、方案及模型进行了仿真分析，，以nDCG算法及tidalTrust算法对应的指标对仿真结果进行评判，以证明相关理论的有效性。
[Abstract]:In order to provide a better user experience, social network applications can publish all kinds of electronic content, such as ordinary text, photos, videos, etc. It involves a lot of personal privacy information, which poses a new challenge to the security and privacy protection of social network applications. Firstly, this paper analyzes the current situation of research on social network application security protection and privacy protection at home and abroad. With the support of relevant theories, an ontology-based computing model for social network user relationship sharing is proposed. In order to extend the one-to-one user relationship and introduce the sharing strategy of the user's group of friends, the blacklist filtering model can be used to the maximum extent. Specifically, this paper first proposes an ontology model of automatically filtering the blacklist of social network based on semantic Web, customizes the existing ontology model of social network, and gives a scheme to construct the ontology model of malicious information of social network. Then, using semantic Web technology, a solution to automatically filter the blacklist of social networks is presented. According to the principle of trust generation in reality, the relevant factors are abstracted and the reasonable classification criteria are given, and the Pearson algorithm is introduced to quantify the internal and external factors of trust between users. Based on the blacklist automatic filtering model, a blacklist sharing model based on user group relationship is proposed to extend the blacklist filtering model. Finally, the relevant theories, schemes and models are simulated and analyzed, and the simulation results are judged by the corresponding indexes of nDCG algorithm and tidalTrust algorithm to prove the validity of the relevant theory.
【学位授予单位】：南京邮电大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.08

【相似文献】