【作者】 赵远萍；

【导师】 舒华英；

【作者基本信息】 北京邮电大学 ， 管理科学与工程， 2010， 博士

【摘要】 随着互联网的发展,即时通信(IM)系统的应用正变得越来越普及,人们在享受IM系统的快捷和便利的同时,也不时被其带来的内容安全问题所困扰。目前IM安全方面的研究主要是针对计算机病毒传播的研究,如：IM系统中蠕虫病毒传播研究等。在IM系统中不仅各种计算机病毒的传播会给人们的生活工作带来危害,各种消息尤其是恶意消息的传播也会对广大网民产生不容忽视的影响。一方面,IM系统可以快速宣传那些具有宣传意义的正面消息,对这类消息的传播应该进行引导和支持；另一方面,虚假信息或是恶意谣言也可以在IM系统中快速扩散,而对于这类消息应该给予控制和管理,并且尽可能在早期采用相应的手段进行监控和制止。因此,与信息安全技术方面的研究相比,信息内容安全治理问题越来越引起社会各界的重视。而目前IM消息的内容安全治理这一领域尚无研究结论,这就使得本文的研究工作变得十分重要。本文正是基于这样的时代和学术背景下进行研究的。本文以信息安全治理的系统化思路作为开展研究的总体思想。从IM消息传播特点和规律的角度,有针对性地提出IM消息传播干预建议。对IM消息传播建模的研究主要分为IM网络拓扑建模和IM消息传播动力学模型分析这样两个部分。具体来讲,本文的主要研究内容及研究成果如下：第一,对IM网络进行拓扑建模分析。在问卷调查得到的数据基础上对IM网络的拓展行为进行分析。基于这些特征制定IM网络的演化机制和模型假设条件。在原有无标度网络演化模型(BA模型)的基础上,结合IM网络的演化机制如：局域优先连接、朋友引荐机制等,对BA模型进行改进。通过与实际网络对比,发现所构建的网络模型能够反映现实IM网络的实际情况。通过仿真分析发现IM网络的节点度分布并不符合幂律分布,网络中具有少量连接和大量连接的节点比例都较小,绝大多数节点的连接数都处在一个范围内。同时,通过对IM网络的其他拓扑性质的研究发现,IM网络是一个具有较短的平均最短路径、较大的聚类系数和清晰的社团结构的复杂网络。第二,对IM系统消息传播模型进行研究。任何一个复杂系统的消息传播模型都包括两个部分：底层的拓扑网络和消息的传播规则。首先通过调研结果总结了IM消息的传播行为特征,引入了适应度参数、边权等概念,对IM消息传播网络进行了建模,并通过仿真发现,该网络的节点度分布具有与IM网络同样的特点。然后在IM消息传播网络的基础上,建立IM消息传播的动力学模型。该模型是在非均匀网络中谣言传播模型的基础上,引入IM用户之间的紧密程度以及消息类型等因素进行改进后建立的。虽然为了研究方便,对模型的参数设置采用了某些简化过程,但是从机理上该模型基本能与实际情况相吻合,可以用于对复杂的消息传播过程进行描述。同时,本文通过实际IM网络的数据进行了模型的验证和分析,发现该模型基本可以描述实际消息的传播趋势。从而得到了一个重要结论,本文所建立的IM消息传播模型可以用于描述IM消息传播过程。第三,基于对IM消息传播趋势的分析提出IM消息的干预机制和建议。以往针对IM系统的安全性研究大多是针对计算机病毒方面的研究,关于消息传播方面的研究基本上都是定性的论述。本文通过建立IM消息传播模型并对仿真结果进行分析,得出了IM消息传播模型的影响因素。这些影响因素包括：接受概率、免疫概率、平均度和网络规模等。并通过不同的参数取值进行仿真实验,得到各个参数对于消息传播趋势的具体影响,为进一步提出针对性的干预建议提供了基础。基于上述工作,本文最后从IM消息传播特点本身以及IM网络拓扑结构两个角度提出了IM系统消息传播的干预建议。并提出IM消息的安全管理是一个持续不断的过程,同时需要全社会的共同努力。本文的研究成果对于更好地了解IM系统的网络拓扑结构、研究其上的消息传播规律以及IM消息传播的干预机制研究等方面都有很强的参考价值。同时,本文采用信息安全治理的系统化思想,按照IM网络拓扑、消息传播、模型验证和干预建议这样几个步骤对IM系统的安全治理进行了研究。这种思想也可以为其他类似的研究提供借鉴。更多还原

【Abstract】 With the development of the Internet, instant messaging (IM) system is becoming more and more popular. While people enjoy the convenient and swift service, they are always plagued by the content security problems of IM system. At present, the safety research for IM system is mainly on computer virus such as IM worm propagation research. Various informations spreading in IM system will produce influences that can not be ignored to people’s daily life and work. On the one hand, the positive news with meaningful information can be quickly diffused in IM system and we should encourage and support such information; on the other hand, there are also many malicious rumors or false information that could rapidly spread in the IM system and we should take appropriate monitoring and suppression measures to control this type of information at the early stage. Comparing with the research of information security technology, the governance of information content security is increasingly aroused the attention. At present, there are no conclusions in the area of content security management for IM system, which makes this study very important. This study is conducted based on the above age and academic background.In this paper, we take the systematic idea of information security management as the basic method to launch the research. Through analysis of information propagation characteristics and laws in IM system, we put forward pertinent suggestions for IM information security governance. The research on information dissemination of IM system can be divided into two parts, which are the analyses of IM topology modeling and information dynamics modeling. The main contents are concluded as follows:Firstly, the topology modeling and characteristics analysis of the IM users’ network are discussed. Based on the specific features of IM users’network from the results of survey, we improve the original scale-free network model (BA model) to match the mechanisms such as local priority connections and friend referral model. Compared with the actual network, we find that the constructed network model can reflect the actual situation of IM users’network.Through the simulation analysis, we find that the IM users’network model doesn’t obey power-law distribution and it has short average path length and high clustering coefficient. Meanwhile, the community structure of the network is also obvious. In this network, there are a small proportion of nodes with little or large connections, and most of the connections are in a specific range.Secondly, the information propagation modeling of the IM system is analysed. Any information dissemination model on complex system includes two parts, which are the underlying topology network and the message transmission rules on it. The transmission rules and characteristics of IM messages are discussed at the beginning of this section based on the results of survey. With the introduction of concepts, such as fitness parameters and side power, we put forward the evolving model for information dissemination network on the IM system. Through the simulation results, we can conclude that the node degree distribution of this network is the same as that of the IM users’ network. Then the dynamic model of IM messages is built on the base of the information dissemination network. With the introduction of the factors meaning tightness of IM users and type of the message, we establish the model based on the rumor propagation model in non-uniform network. Although we set the parameters of the model using a simplified process, the mechanism on the model can match with the actual situation and we can describe the complex process of information dissemination clearly. Meanwhile, through empirical study with the actual IM network in this paper, we find the model can describe the propagation trend of the IM messages probably. Thus we can make an important conclusion that this model which is constructed in this paper can be used to analyze the propagation situation of the IM messages.Thirdly, through the analysis of the IM messages dissemination trend, we propose the intervention suggestions of content security for IM system. In the past, the studies on IM system security have focused on computer viruses, and researches on information dissemination are basically qualitative discussion. Aiming at finding the factors of IM message transmission, we establish the IM message propagation model and analyze the simulation results. These factors include acceptance probability, rejection probability, the average degree and the network scale. Through different experimental parameters, we obtain the specific impact to the trend of information dissemination, which provide a basis for the study of further specific intervention recommendations. Based on the above work, this article puts forth the intervention suggestions for IM message from two aspects, which are the structure of IM network and the characteristics of message transmission. This intervention management is a continuous process, which needs to adopt a systematic approach for monitoring and management. And it also needs the joint efforts of the whole society.The achievements in the paper have significant roles to understand the network topology of IM system, the propagation characteristics of the message and the intervention mechanisms. Meanwhile, this paper obeys to the standard steps as network topology, information propagation, concrete evidence research and policy recommendations. This thinking is good to other similar research.更多还原