腾讯微博的内容生成模式分析

随着Web 2.0 时代网络技术的快速发展,社交类网站用户大规模增加。该文选取腾讯微博近两万名用户,抓取了他们所有的微博数据,对腾讯微博的用户内容生成模式进行分析和研究。我们从微博用户贡献分析、基于时间的用户活跃度分析以及微博影响三个角度出发,对微博的数量、微博的原创与转发、微博发布的周模式与日模式、微博转发影响力以及对影响微博转发的因素进行研究。总结出微博用户内容生成的一些特点,如用户内容贡献呈现一种“90-10”规则,不同类型的用户有着不同的“微博风格”,微博用户每日微博发布数有着明显的周模式与日模式等。相关分析结论对于进一步深化研究微博的用户内容生成模式具有一定参考意义。     

基于链路预测的微博用户关系分析

随着以微博为代表的在线社交网站的发展,微博用户之间形成了复杂的社会网络。针对微博社会网络,研究了影响微博用户之间关系形成的各种因素,提出了基于链路预测的微博用户关系分析模型。首先分析了网络结构特征在微博社会网络中的作用,同时针对微博社会网络的特点,引入微博属性特征,构造基于随机森林的链路预测模型,并将模型应用于新浪微博用户数据集,进行微博用户关系的训练预测,通过比较引入微博属性特征前后的预测性能以及特征的重要性分布,分析了各类特征对微博用户关系形成的影响,揭示了除传统的网络结构特征外,微博属性特征对微博用户关系的形成具有重要的影响力。     

基于LDA模型的微博话题发现技术研究

微博中存在着数以亿计的用户,这些用户每天发布大量的信息。这些海量的微博信息给热点话题发现提出了严峻的挑战。应用LDA(Latent Dirichlet Allocation)模型对微博中隐含的话题进行建模,利用话题间的共享词汇将话题构成一个无向加权图,并通过PageRank算法将话题进行排名。实验结果表明,排名后返回给用户的话题的准确性明显高于未排名的结果。     

Exploring the effect of user engagement in online brand communities: Evidence from Twitter

Social media such as forums, blogs and microblogs has been increasingly used for public information sharing and opinions exchange nowadays. It has changed the way how online community interacts and somehow has led to a new trend of engagement for online retailers especially on microblogging websites such as Twitter. In this study, we investigated the impact of online retailers' engagement with the online brand communities on users' perception of brand image and service. Firstly, we analysed the overall sentiment trends of different brands and the patterns of engagement between companies and customers using the collected tweets posted on a popular social media platform, Twitter. Then, we studied how different types of engagements affect customer sentiments. Our analysis shows that engagement has an effect on sentiments that associate with brand image, perception and customer service of the online retailers. Our findings indicate that the level, length, type and attitude of retailers' engagement with social media users have a significant impact on their sentiments. Based on our results, we derived several important managerial and practical implications.     

基于行为特征分析的微博恶意用户识别

近年来,社交网络数据挖掘作为物理网络空间数据挖掘的一大热点,目前在用户行为分析、兴趣识别、产品推荐等方面都取得了令人可喜的成果。随着社交网络商业契机的到来,出现了很多恶意用户及恶意行为,给数据挖掘的效果产生了极大的影响。基于此,提出基于用户行为特征分析的恶意用户识别方法,该方法引入主成分分析方法对微博网络用户行为数据进行挖掘,对各维度特征的权重进行排序,选取前六维主成分特征可以有效识别恶意用户,主成分特征之间拟合出的新特征也能提升系统的识别性能。实验结果表明,引入的方法对微博用户特征进行了有效的排序,很好地识别出了微博社交网络中的恶意用户,为其他方向的社交网络数据挖掘提供了良好的数据清洗技术。     

Professional or interactive: CEOs’ image strategies in the microblogging context

Microblogging platforms are gaining popularity among corporations and their top management in recent years. Although microblogging services like Twitter and Sina Weibo are now prevalently used for managing CEOs’ images and public relations, few studies have examined the effects of these practices on the loyalty of target audiences. This study examined the effect of CEO image strategy on follower loyalty in the microblogging context. Based on the self-presentation theory, four types of CEO image strategies were identified, namely the Expert, Friend, Textbook, and Daybook strategies. These categories were identified based on the levels of interactivity and professionalism of the CEOs on their microblogs. An online survey was used to collect data from microblog users, that is, the CEOs’ followers. The results showed that CEO image strategy influences follower loyalty in the microblogging context and that Chinese microblog users are fondest of CEOs who present themselves as experts rather than as friends (H1). The results also showed that usage orientation moderates the effect of CEO image on follower loyalty (H2) and that goal orientation positively influences CEOs with a highly professional image.     

Microblogs in Higher Education – A chance to facilitate informal and process-oriented learning?

Microblogging is one of the latest Web 2.0 technologies. The key elements are online communication using 140 characters and the fact that it involves “following” anyone. There has been a great deal of excitement about this in recent months. This paper reports on a research study that was carried out on the use of a microblogging platform for process-oriented learning in Higher Education. Students of the University of Applied Sciences of Upper Austria used the tool throughout their course. All postings were carefully tracked, examined and analyzed in order to explore the possibilities offered by microblogging in education. It can be concluded that microblogging should be seen as a completely new form of communication that can support informal learning beyond classrooms.     

基于协商理念的微博和微视频英语教学研究

大班教学成为当前国内高校公共外语教学的一种普遍现象,由此带来的课堂师生互动减少,教学效率不高的问题,亟待我们解决。信息化时代,微博碎片式的网络互动交流模式可以弥补大学英语教学课堂的单一模式,能为师生创造课后外语交流的平台和情境,促进课内课外形成整体,从而激发学生学习外语的热情,发挥其学习主观能动性,促进大学外语教学的进步。     

分级意见领袖下的微博信息传播与数学建模

为了有效地分析意见领袖在微博在线社交网络传播中所起的作用及微博信息传播的生命周期,提出了一种OLL分级意见领袖模型。首先,爬取微博数据,并进行数据统计分析;其次,将传播力构建为转发数、活跃度与粉丝量3个因素相关的函数,并建立一种基于层次分析的权重计算方法;最后, 利用计算的传播力并结合OLL模型,对意见领袖传播作用及微博生命周期进行了实验分析。仿真结果表明:意见领袖在微博信息传播中有很强的传播效应,通过计算得出OLL模型仿真与3组真实数据的误差分别为9.6%,13.4%,4.5%,从而证明了所提OLL模型对于分析意见领袖在微博信息传播中的生命周期是合理且有效的。