基于复杂网络的微博信息传播规律研究

发布时间：2018-05-20 04:13

本文选题：微博信息传播 + 信息扩散级联　；参考：《哈尔滨工业大学》2014年硕士论文

【摘要】：现在是一个信息化的时代，我们每天都接触大量的信息，随着网络媒体的不断推进和演化，微博成为一个重要的信息交流平台，，改变了传统的信息交互方式。人们不但是信息的传播者，而且也是信息的制造者，信息的内容更加多样化，信息传播的时间也更加碎片化。对微博信息传播的规律进行挖掘和发现，可以使得我们深刻理解人们信息交互的本质，为微博营销和舆情监控等提供了理论依据。本文结合复杂网络的理论，对微博环境的信息传播进行了整体的分析，包括对微博信息传播网络的网络结构、微博信息传播的特征和演化规律的研究，对微博信息传播路径和时间特性的分析，以及对用户信息传播能力的度量。本文用用户之间的转发行为来刻画微博环境中的信息传播，首先随机选取了45770名新浪微博用户，选取了他们截止到2013年9月30日为止最新的100条微博，然后从每个用户的100条微博中提取转发记录；对原始数据进行清洗之后构建了用户转发网络，同时为了进行对比，本文也基于用户之间的关注关系构建了用户关注网络；分析了两个网络在拓扑结构方面的特点，并对信息传播网络中信息传播的特征和信息传播网络的拓扑指标随样本量大小的演化规律进行了分析。其次，进一步分析了微博中信息传播路径的分布及其与用户粉丝数的关系，并重点分析了微博信息传播的时间规律，包括传播时间和时间延迟两个方面。最后，综合了表征微博用户自身属性的三个指标和表征用户在信息传播网络中传播能力的七个指标，以主成分分析为理论依据，构建了度量用户影响力的数学模型，并给出了信息传播能力排名前20的用户。
[Abstract]:Now is an information age, we come into contact with a lot of information every day, with the continuous promotion and evolution of network media, Weibo has become an important information exchange platform, which has changed the traditional way of information exchange. People are not only the disseminators of information, but also the makers of information. The content of information is more diversified, and the time of information dissemination is more fragmented. Mining and discovering the rules of Weibo information dissemination can make us understand the essence of information exchange deeply and provide theoretical basis for Weibo marketing and public opinion monitoring and so on. Based on the theory of complex network, this paper analyzes the information transmission in Weibo environment, including the network structure of Weibo information communication network, the characteristics and evolution law of Weibo information transmission. This paper analyzes the characteristics of Weibo information transmission path and time, and measures the ability of user information transmission. In this paper, the forwarding behavior between users is used to describe the spread of information in the Weibo environment. First, 45770 Sina Weibo users are randomly selected, and their latest 100 Weibo as of Sept. 30, 2013, are selected. Then the forwarding records are extracted from each user's 100 Weibo. After cleaning the original data, the user forwarding network is constructed. In order to compare, this paper also constructs the user concern network based on the concern relationship between the users. In this paper, the characteristics of the two networks in topology structure are analyzed, and the characteristics of information transmission in the information dissemination network and the evolution law of the topological index of the information dissemination network with the sample size are analyzed. Secondly, the distribution of information transmission path in Weibo and its relationship with the number of users are analyzed, and the time law of Weibo information transmission is analyzed, including the time and time delay. Finally, three indexes representing the attributes of Weibo users and seven indicators representing the ability of users to spread in the information dissemination network are synthesized. Based on principal component analysis (PCA), a mathematical model is constructed to measure the influence of users. And gives the information dissemination ability ranking top 20 users.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.092;O157.5

【参考文献】