基于社会网络理论的恐怖组织隐蔽网络研究

发布时间：2018-06-22 07:10

本文选题：恐怖主义 + 社会网络分析　；参考：《国防科学技术大学》2010年硕士论文

【摘要】：社会网络分析(Social Networks Analysis,简称SNA),又称为网络分析,结构分析或人际关系研究,是指一种以社会行动者之间的互动研究为基础的结构性方法,它以图论数学为基础,用点来表示社会行动者,用点和点间的连线来表示行动者间的某种社会关系,从而对这些社会关系可以进行量化分析。社会网络分析方法在70年代以来得到了快速的发展,被广泛应用于几乎所有的社会学领域。在国家安全领域也开始引起广泛关注,并在“9.11”后成为热点。将社会网络分析应用于恐怖网络的研究,即用节点表示恐怖分子和与之有关的技能、资源、任务、地点等要素,用连线来表示恐怖分子与各要素之间存在关系,其中恐怖分子间的关系及其形成的社交网络是最为重要的分析内容之一。这些社会关系有很多种,如恐怖分子间的血亲关系、族群关系、宗派关系、友谊关系、有共同训练或工作的经历等。作为一种研究范式,社会网络比传统个体主义的研究途径更具社会学解释力,它直接针对社会结构的模式化的关系本质,具有传统主流统计方法没有的优势;作为一种数据挖掘技术,社会网络分析能处理大量恐怖情报数据,从中挖掘出有用的信息,具有防范和打击恐怖活动的现实应用价值。本文从社会网络理论的视角分析恐怖组织及其行为,首先介绍社会网络分析的发展轨迹及一般方法,阐述了恐怖网络分析的主要内容和方法论特征,并总结新恐怖主义的发展、网络视野下的恐怖主义及其组织结构模型、发展演变的原因和特点,以此阐明社会网络分析与传统研究途径的相比的优势所在;介绍了对恐怖分子社会网络的构建与分析的基本方法,提出基于中心度、子群参与指数和指控结构的网络个体成员分析方法,从中可分析成员在网络中扮演的角色及作用,及确定网络的脆弱点所在,对网络关键成员的定位及网络行为模式的测量使用美国大使馆为例作了检验分析。本文最后尝试对东突恐怖组织进行网络视角的有益探索,通过搜集主流媒体网络与东突有关的恐怖事件资料110个网络文本,并以此为数据来源对东突恐怖网络进行构建和分析,文中主要工作使用的是AutoMap网络文本分析工具及UCINET和ORA (Organization Risk Analyzer )为网络可视化分析工具。研究表明,社会网络分析方法在恐怖组织隐蔽网络分析中的应用具有传统个体属性统计分析方法所没有的优势,恐怖分子的社会关系及其在实施恐怖活动时产生的关联数据隐藏着可用于侦查发现其组织活动或预测其恐怖活动规律的信息。通过将收集的恐怖分子各种关系数据输入网络分析工具,可以将恐怖组织隐蔽的社会网络视觉化为网络图模型,密度测量可分析恐怖组织成员间的交流量并捕捉其行为特性,凝聚子群和结构等价性等的测量则帮助发现恐怖组织中的内部帮派,进而能分析恐怖组织隐蔽网络的特征,为了解恐怖组织当前所处发展阶段及运作模式提供思考,从中并可观察恐怖组织的发展、演变过程。在个体层次的分析上,本研究通过结合结构等价分析、人际交互流向的指控结构、网络中心度量化及派系参与方法,挖掘出了恐怖组织隐蔽网络核心人物,分析其扮演的角色并找出了瓦解网络的脆弱点。通过网络文本分析的方法较快地处理了资料来源,从110个网络文本中抽取出与研究有关的关键实体及其关联数据,进而构建出了东突恐怖网络图模型,分析发现基于语义分析获得的东突恐怖二模网络与典型的随机网络、核心-边缘网络及蜂窝网络均不相同等特征,但其子网络如一模社会网络及组织关联网络具有典型网络的特征;其核心网络结构明显,通过中心度测量方法定位的网络核心成员得到了验证,在稀疏的网络密度中,存在大量接近性强的成员,且相互关联的组织名目繁多,亦即其各群组间的结构不具强等级关系,虽然东突成员大多联系紧密,但以隶属于不同组织的方式加以表现。
[Abstract]:Social Networks Analysis (referred to as SNA), also known as network analysis, structural analysis, or interpersonal relationship, refers to a structural method based on the interaction between social actors. It is based on graph mathematics to represent social actors with points and points between actors. Social relations can be quantified and analyzed. Social network analysis has been developed rapidly since 70s and is widely used in almost all fields of sociology. It has also attracted wide attention in the field of national security and became a hot spot after "9.11". Social network analysis is applied to the social network analysis. The study of terrorist networks, which uses nodes to express terrorists and related skills, resources, tasks, and locations, uses links to express the relationship between terrorists and various elements, among which the relationship between terrorists and the social network formed is one of the most important analysis. There are many kinds of social relations, such as fear. As a research paradigm, the social network is more sociological than the traditional individualist approach, which directly aims at the nature of the pattern of the social structure and has no advantages in the traditional mainstream statistical methods. As a data mining technology, social network analysis can handle a large number of terrorist information data, excavate useful information from it, and have practical application value to prevent and combat terrorist activities.
This paper analyzes the terrorist organization and its behavior from the perspective of social network theory. First, it introduces the development track and general methods of social network analysis, expounds the main contents and methodological features of the terrorist network analysis, and summarizes the development of the new terrorism, the terrorism and its organizational structure model in the network field of vision, and the reasons for its development and evolution. In order to clarify the advantages of social network analysis compared with the traditional research approach, this paper introduces the basic methods for the construction and analysis of the social network of terrorists, and proposes a network individual member analysis method based on the center degree, the subgroup participation index and the accusation structure, from which the role and the role of the members in the network can be analyzed. Using and determining the vulnerabilities of the network, the location of the key members of the network and the measurement of network behavior patterns are tested and analyzed by the American embassy. Finally, this paper tries to explore the network perspective of the East Turkistan terrorist organization and collect the 110 network texts of the terrorist information related to the East Turkistan in the mainstream media network. This is the data source for the construction and analysis of the East Turkistan terrorist network. The main work used in this paper is AutoMap network text analysis tools and UCINET and ORA (Organization Risk Analyzer) as a network visualization analysis tool.
The research shows that the application of social network analysis method in the analysis of hidden network in terrorist organizations has the advantages of the traditional individual attribute statistical analysis methods. The social relations of terrorists and the related data produced in the implementation of terrorist activities can be hidden to detect their organizational activities or predict their terrorist activities. Information. Through the input of various relational data of the terrorists to network analysis tools, the hidden social network of terrorist organizations can be visualized as a network graph model. Density measurement can analyze the amount of communication among members of terrorist organizations and capture their behavior characteristics, and the measurement of aggregation subgroups and structure equivalence can help to find terrorist organizations. In order to understand the current development stage and mode of operation of terrorist organizations, the internal gangs of internal gangs can analyze the characteristics of terrorist organizations' hidden networks, and observe the development and evolution process of terrorist organizations. Collaterals center degree quantification and factional participation method, excavate the core character of the hidden network of terrorist organization, analyze its role and find out the fragility of the disintegration network. Through the method of network text analysis, the data sources are processed quickly, and the key entities and related data related to the research are extracted from the 110 network texts. The model of East Turkistan terrorist network graph is built, and it is found that the two mode network of East Turkistan terrorism based on semantic analysis is different from the typical random network, the core edge network and the cellular network are different, but the sub network, such as one mode social network and the organization association network, has the characteristics of typical network; its core network structure is obvious. The core members of the network are verified by the centrality measurement method. In the sparse network density, there are a large number of close members and interrelated organizations, that is, the structure of each group has no strong hierarchical relationship. Although the members of the East Turkistan are closely connected, they are added to different organizations. To show.
【学位授予单位】：国防科学技术大学
【学位级别】：硕士
【学位授予年份】：2010
【分类号】：D815.5

【参考文献】