BBS用户行为分析

发布时间：2018-02-13 03:40

本文关键词： BBS论坛复杂网络用户分类互动行为交互特性　出处：《首都师范大学》2014年硕士论文　论文类型：学位论文

【摘要】：随着互联网和Web2.0的飞速发展,许多供人们交流和沟通的虚拟空间也随之产生,例如博客、BBS论坛等。在BBS论坛中,用户之间通过发帖和回帖的形式建立联系,形成虚拟社区中一种新的互动形式。由于虚拟社区与现实社区存在不同特性,国内外不少研究者都对BBS用户行为进行了相关研究。目前对BBS用户行为的研究主要集中在两个方面：一是BBS用户行为宏观分析,即构建BBS用户回复网络,着重研究BBS用户回复网络的小世界特性和无标度特性；二是BBS用户行为微观分析,通过分析用户之间的微观互动行为,掌握BBS用户行为的特征和用户之间互动形式。本论文以新浪BBS论坛数据构建的回复网络为研究对象,利用复杂网络理论分析BBS用户回复网络的拓扑结构特性,研究BBS用户回复网络的小世界特性和无标度特性。并通过定量分析BBS用户行为,明确用户在BBS社区中的作用及互动行为特点,对保障BBS社区健康发展有重要意义。本文的研究内容如下： (1)根据用户之间的回复关系构建BBS用户回复网络,利用pajek软件实现回复网络的可视化。基于复杂网络的基本理论,根据聚类系数、平均路径长度、度分布指标等特征参数,研究BBS用户回复网络的拓扑结构。 (2)定量分析BBS用户行为,通过统计用户发帖量和用户ID量,发现用户行为特征。 (3)选取BBS用户发帖总数、发主题帖数、获回帖数、用户回帖数和发表主题帖被纳入精华帖的总数作为描述用户行为的分类指标,利用聚类算法实现BBS社区用户的分类,根据分类结果将社区成员分为领袖型、回应型、浏览型和实力型用户,并根据成员相似度和成员关联强度分析用户的行为特征和交互特点。 (4)引入主题树概念,利用树形论坛的特殊结构构建用户之间的回复关系,提出主题深度、主题广度等参数描述用户的互动行为。分析主题广度系数W和主题综合深度系数D综合评价主题的互动情况,并根据互动特征将用户的互动行为分为五类,对BBS社区信息管理起到积极的指导意义。
[Abstract]:With the rapid development of the Internet and Web2.0, there are many virtual spaces for people to communicate and communicate, such as blogs and Web2.0 forums. In BBS forums, users establish contact by posting and responding to each other. Because there are different characteristics between virtual community and real community, many researchers at home and abroad have done some research on BBS user behavior. At present, the research on BBS user behavior mainly focuses on two aspects: first, the macro analysis of BBS user behavior, that is, the construction of BBS user response network, focusing on the small-world and scale-free characteristics of BBS user response network; Secondly, the microcosmic analysis of BBS user behavior, through analyzing the microcosmic interaction behavior between users, grasps the characteristics of BBS user behavior and the interaction form between users. In this paper, we use the complex network theory to analyze the topology characteristics of BBS user response network, which is based on the data of Sina BBS forum. This paper studies the small-world and scale-free characteristics of BBS user response network, and through the quantitative analysis of BBS user behavior, clarifies the role and interactive behavior of users in BBS community, which is of great significance to ensure the healthy development of BBS community. The contents of this paper are as follows:. Based on the basic theory of complex network, based on the clustering coefficient, the average path length, the degree distribution index and other characteristic parameters, the BBS user response network is constructed according to the response relation between the users and the pajek software. The topology of BBS user response network is studied. Second, quantitative analysis of BBS user behavior, by counting the number of user posts and user ID, found the user behavior characteristics. 3) selecting the total number of BBS users' posts, the number of theme posts, the number of replies received, the total number of users' replies and the total number of posts published into the essence of posts as classification indicators to describe the user's behavior, and using clustering algorithm to realize the classification of users in BBS community. According to the classification results, the community members are divided into leadership type, response type, browsing type and strength type, and the behavior and interaction characteristics of users are analyzed according to the similarity of members and the intensity of association. (4) introducing the concept of theme tree, using the special structure of the tree forum to construct the response relationship between users, and putting forward the depth of the topic. The interactive behavior of users is described by such parameters as theme breadth, and the interaction of theme breadth coefficient W and theme comprehensive depth coefficient D is analyzed, and the user's interactive behavior is divided into five categories according to the interactive characteristics. BBS community information management plays a positive guiding significance.
【学位授予单位】：首都师范大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.09;O157.5

【参考文献】