面向协作学习的社会网络结构分析与用户行为建模
发布时间：2018-11-24 19:45
【摘要】：作为新兴的社会化软件,博客,掀起了一场互联网世界的新革命,为人们带来了一种新的交流方式。学术界对它也是日益关注,对它的研究领域也已经涉及到新闻学、传播学、计算机网络技术和社会学等学科。通过对博客在中国发展现状的实证研究,为社交网络平台以及其他领域的发展提供数据依据。本文的研究对象是社交网络,以及如何促进博客间的交流。以理论为依据应用统计方法和社交网络分析方法对博客网络进行分析和描述。通过社交网络中存在的用户行为,对其进行详尽的挖掘和分析,探寻影响这些用户行为的关键性因素。同时,本文尝试建立准确的模型来刻画用户的行为,从而为实际的应用需求服务。在本文阐述的研究工作不光注重懫用合理、有效的理论方法,更看重实验结果和相关结论对实际的应用场合能带来的促进作用。关于本文的具体研究内容,主要包含五个章节。第一章,主要是从当前社交网络平台的背景出发阐述该课题的研究意义及应用领域,介绍了本论文准备解决的问题以及解决这些问题所用的相关方法;第二章,主要介绍了关于社交网络方面的一些基本概念和理论,首先介绍了社交网络分析的相关知识,比如中心度的一些概念等;紧接着对网络传播动力学展开讨论,重点介绍了几种重要的传播模型;接下来对微博用户的特征与类型进行了分析;最后对于用户行为进行了详细研究并介绍了常用的用户行为模型;第三章,本章以用户为中心,以微博中的热点话题为例,进行了在社交网络中的不同的用户角色研究,通过考虑一个在线社交网络的微博系统,提出了基于关键词的隐式链接和用户关系网络模型;第四章,本章主要的研究内容是对用户的行为规律进行总结,并对用户行为、社交网络中信息的传播机制进行了深入的分析,提出了社交网络中信息的传播模型;第五章,本章节先对协作学习做了基本介绍,然后借助社交网络分析工具,对模糊多属性决策问题在协作学习的方面,进行了完全评价以及改进了属性权重信息的研究,再根据TOPSIS法的基本思想,计算出了每个方案之间的Hamming距离正理想解与负理想解,然后,根据加权Hamming距离,计算了相对贴近度到正理想方案的排名,最后,给出一个具体例子社交网络分析在协作学习的评价研究。论文以研究社交网络结构为切入点,构建了基于兴趣衰减和自适应兴趣驱动模型、非齐次泊松模型的在线热点话题关注模型,进而研究社交网络对于协作学习的影响。模型对于学习的过程起到了关键性的作用,对于模型选择的探索提供了便利的条件。
[Abstract]:As a new social software, blog has set off a new revolution in the Internet world and brought a new way of communication for people. Academic circles also pay more and more attention to it, and its research fields have been related to journalism, communication, computer network technology and sociology and so on. Through the empirical research on the status quo of blog development in China, it provides the data basis for the development of social network platform and other fields. The research object of this paper is social network and how to promote the communication between blogs. Based on the theory, the blog network is analyzed and described by using statistical method and social network analysis method. Through the user behavior in the social network, the author excavates and analyzes the user behavior in detail, and explores the key factors that affect the user behavior. At the same time, this paper attempts to establish an accurate model to describe the behavior of users, so as to serve the actual application requirements. The research work described in this paper not only focuses on the rational and effective theoretical methods, but also on the promotive effect of the experimental results and related conclusions on the practical applications. The specific research content of this paper mainly includes five chapters. In the first chapter, from the background of the current social network platform, the research significance and application field of this topic are expounded, and the problems to be solved in this paper and the relevant methods to solve these problems are introduced. The second chapter mainly introduces some basic concepts and theories about social network. Firstly, it introduces the relevant knowledge of social network analysis, such as some concepts of centrality. Secondly, it discusses the dynamics of network communication, introduces several important communication models, and analyzes the characteristics and types of Weibo users. Finally, the user behavior is studied in detail and the commonly used user behavior models are introduced. In the third chapter, this chapter takes the user as the center and the hot topic in Weibo as an example to study the different user roles in the social network. An implicit link and user relationship network model based on keywords is proposed. In the fourth chapter, the main research content of this chapter is to summarize the behavior of users, analyze the behavior of users and the mechanism of information dissemination in social networks, and put forward the model of information dissemination in social networks. In the fifth chapter, we first introduce the cooperative learning, and then, with the help of social network analysis tools, we evaluate the fuzzy multi-attribute decision making problem in the aspect of collaborative learning and improve the research of attribute weight information. Then, according to the basic idea of TOPSIS method, the positive and negative ideal solutions of Hamming distance between each scheme are calculated. Then, according to the weighted Hamming distance, the relative closeness degree to the ranking of positive ideal schemes is calculated. A concrete example of the evaluation of social network analysis in collaborative learning is given. Based on the research of social network structure, this paper constructs an online hot topic focus model based on interest attenuation and adaptive interest driven model and non-homogeneous Poisson model, and then studies the influence of social network on collaborative learning. The model plays a key role in the learning process and provides a convenient condition for the exploration of model selection.
【学位授予单位】：吉林大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：TP393.092
本文编号：2354815
[Abstract]:As a new social software, blog has set off a new revolution in the Internet world and brought a new way of communication for people. Academic circles also pay more and more attention to it, and its research fields have been related to journalism, communication, computer network technology and sociology and so on. Through the empirical research on the status quo of blog development in China, it provides the data basis for the development of social network platform and other fields. The research object of this paper is social network and how to promote the communication between blogs. Based on the theory, the blog network is analyzed and described by using statistical method and social network analysis method. Through the user behavior in the social network, the author excavates and analyzes the user behavior in detail, and explores the key factors that affect the user behavior. At the same time, this paper attempts to establish an accurate model to describe the behavior of users, so as to serve the actual application requirements. The research work described in this paper not only focuses on the rational and effective theoretical methods, but also on the promotive effect of the experimental results and related conclusions on the practical applications. The specific research content of this paper mainly includes five chapters. In the first chapter, from the background of the current social network platform, the research significance and application field of this topic are expounded, and the problems to be solved in this paper and the relevant methods to solve these problems are introduced. The second chapter mainly introduces some basic concepts and theories about social network. Firstly, it introduces the relevant knowledge of social network analysis, such as some concepts of centrality. Secondly, it discusses the dynamics of network communication, introduces several important communication models, and analyzes the characteristics and types of Weibo users. Finally, the user behavior is studied in detail and the commonly used user behavior models are introduced. In the third chapter, this chapter takes the user as the center and the hot topic in Weibo as an example to study the different user roles in the social network. An implicit link and user relationship network model based on keywords is proposed. In the fourth chapter, the main research content of this chapter is to summarize the behavior of users, analyze the behavior of users and the mechanism of information dissemination in social networks, and put forward the model of information dissemination in social networks. In the fifth chapter, we first introduce the cooperative learning, and then, with the help of social network analysis tools, we evaluate the fuzzy multi-attribute decision making problem in the aspect of collaborative learning and improve the research of attribute weight information. Then, according to the basic idea of TOPSIS method, the positive and negative ideal solutions of Hamming distance between each scheme are calculated. Then, according to the weighted Hamming distance, the relative closeness degree to the ranking of positive ideal schemes is calculated. A concrete example of the evaluation of social network analysis in collaborative learning is given. Based on the research of social network structure, this paper constructs an online hot topic focus model based on interest attenuation and adaptive interest driven model and non-homogeneous Poisson model, and then studies the influence of social network on collaborative learning. The model plays a key role in the learning process and provides a convenient condition for the exploration of model selection.
【学位授予单位】：吉林大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：TP393.092
【参考文献】
相关期刊论文 前5条
1 曾繁旭;黄广生;;网络意见领袖社区的构成、联动及其政策影响:以微博为例[J];开放时代;2012年04期
2 袁毅;杨成明;;微博客用户信息交流过程中形成的不同社会网络及其关系实证研究[J];图书情报工作;2011年12期
3 赵文兵;朱庆华;吴克文;黄奇;;微博客用户特性及动机分析——以和讯财经微博为例[J];现代图书情报技术;2011年02期
4 杨博;刘大有;金弟;马海宾;;复杂网络聚类方法[J];软件学报;2009年01期
5 周文坤,武振业,鞠廷英;多目标群体决策的一种综合集成方法[J];西南交通大学学报;2001年01期
，本文编号：2354815
本文链接：https://www.wllwen.com/shoufeilunwen/xxkjbs/2354815.html
