【作者】 王辉；

【导师】 韩江洪；

【作者基本信息】 合肥工业大学 ， 计算机应用技术， 2013， 博士

【摘要】 谣言传播是一个社会生活中的常见现象。在互联网时代,由于在线社交网络这种新型媒体大量出现,人们的表达方式越来越自由,人们传播信息的代价越来越小。这也导致在虚拟的社会网络上传播的信息中有很多是未经证实的传闻,也就是谣言。很多谣言对社会稳定的危害是非常大的,因此研究在线社交网络中的谣言传播特点,掌握其传播规律,从而进行有效的免疫和控制,是一个非常值得关注的问题。复杂网络上的传播动力学问题是复杂网络研究的一个重要分支,在过去的20年中,复杂网络中的流行病传播、计算机病毒传播以及谣言传播等方向都有很多重要的研究成果。本文围绕着在新型的在线社交网络中谣言传播的特点和规律来展开研究。研究的内容主要涉及：(1)在线社交网络中谣言传播的特点与传统谣言传播模型的特点相比有哪些不同,如何用新的方法来划分在线社交网络中人群的种类,用复杂网络研究的方法来研究在线社交网络中人们的谣言传播动力学行为。(2)社会加强现象是现实的谣言传播中的常见现象,在新型在线社交网络上这个现象尤其突出,传统的谣言传播模型并没有对这个现象进行描述,如何在新的传播模型中对该现象进行定义和描述。(3)移动互联网络的兴起,让移动社交网络成为新的研究热点,在移动社交网络中,人们的谣言传播行为和社会加强效应有哪些新的特点,如何用实验来进行仿真。本文的研究内容和主要创新点如下：(1)在线社交网络中,人群应该重新进行划分,人们的谣言传播动力学行为应该重新定义。本文作者提出了CSR(Credulous-Spreader-Rationals)模型,将人群重新划分为轻信者、传播者和理性人三类,并且重新定义了三者的传播行为,使之更符合社交网络中的谣言传播特点。并用平均场理论对动力学行为进行了理论的分析和推导。数值仿真实验表明,CSR模型在小世界网络中的传播效果比传统模型要好。(2)社会加强效应在近期很多学者都进行了研究,为了进一步对社会加强效应进行深入研究,本文提出了双向社会加强的概念,并且给出了数学模型。并用多主体仿真平台对考虑了社会加强效应的CSR模型进行了多主体仿真研究。实验表明,考虑了社会加强效应的模型比传统模型传播更快,范围更广。(3)移动社交网络中的谣言传播有很多新的特点。本文在考虑了移动社交网络传播特点的基础上对CSR模型进行了改进,并对人群的传播动力学行为进行了重新的定义和描述且做了理论上的分析和推理。在接受概率的数学模型中,增加考虑了个人接受阈值对谣言传播的影响。多主体仿真实验的结果表明,新的模型在匀质网络中的传播速度比CSR模型和传统模型更快,传播范围更广。但是在异质网络中,3者传播的效果并没有明显的区别。同时,仿真实验表明模型具有初值敏感性的特点。更多还原

【Abstract】 Rumor spreading is a common phenomenon in social life. In information age, because of the popularity of a new kind of media---the online social networks, people express their opinions more freely, and the cost of rumor spreading in online social networks becomes lower, which leads to much false information, known as rumor, spreading in online social networks. Rumor spreading always comes with upset feelings and threatens social stability. Therefore, it’s extremely important to study the characteristics of rumor spreading in online social networks so as to prevent and control it effectively.The dynamic of information spreading on complex networks is an important branch of complex networks study. In the past twenty years, there have been many significant achievements in epidemics, computer virus propagation and rumor spreading. This study focuses on the characteristics of rumor spreading in online social networks, the main content includes:(1)what are the differences between the characteristics of rumor spreading in online social networks and the characteristics of conventional rumor spreading pattern, how to divide people in online social networks in a new way, and to study people’s behavior in online social networks with the method of complex networks study.(2)social reinforcement is a common and apparent phenomenon in rumor spreading, which hasn’t been explained in conventional rumor spreading pattern, so the paper will define and describe the phenomenon in the new rumor spreading pattern.(3) the study of mobile social networks becomes a hot topic because of the popularity of mobile online networks, so the paper studies the new characteristics of rumor spreading behavior, and how to study in simulation experiment.The main contributions are as follows:1) In this research, a new rumor spreading model for rumor propagation in online social networks, Credulous-Spreader-Rationals(CSR) model, is developed. Analytically, a mean field theory is worked out by considering the influence of network topological structure. Under certain conditions, rumor spreads faster in CSR model than other rumor spreading models in online social networks.2) Classic rumor spreading model overlooked the social reinforcement influence. But this influence always exists in real society or online social networks. Recently, researchers started to take this influence into account. There is a clearly distinguished feature between "social reinforcement" in this study and the rest of papers. We assume social reinforcement contains positive and negative aspects, namely, if the Credulous doesn’t accept the rumor when he meets Spreader, acceptant probability will be increased as positive social reinforcement, it will be easier for him to accept the rumor at next time. However, if the Credulous meets Rationals after spreading process starts, the rumor acceptant probability of this Credulous will be declined because of the negative social reinforcement effects. In CSR model, negative and positive social reinforcements are considered in the acceptant probability model.3) In this paper present, we propose an improved CSR model for rumor spreading in Mobile social networks.The dynamic equation of rumor spreading is modified to be suitable for user’s habit in Mobile social networks. In the acceptant probability model, negative and positive social reinforcements are considered. Furthermore, the people’s Accepting threshold for rumor accepting is taken into account. Analytically, a mean field theory is worked out by considering the influence of network topological structure as homogeneous. Under certain conditions, rumor spreads faster and wider in new model than CSR rumor spreading model in homogeneous networks. Meanwhile, the Multi-agent simulation results indicate that the information spreading process is sensitive dependent on initial conditions.更多还原