社交网络中的社团结构挖掘

发布时间：2017-12-27 15:37

本文关键词：社交网络中的社团结构挖掘　出处：《中文信息学报》2014年01期 　论文类型：期刊论文

【摘要】：社交网络已经成为现代人们在线交流并交换信息的重要途径之一。以国内的人人网为例,大量的年轻人,尤其是学生,以此为平台,相互讨论感兴趣的话题。人与人之间因为学习关系、工作关系、共同的兴趣等诸多因素关联起来;以大学生交流为主体的社交网则更有可能因为在相同院、系、所而关联在一起,从而呈现出社团结构。该文以人人网的真实数据,使用CNM算法来验证这一假设;同时,还利用社会网络的结构知识对CNM算法作了改进,提高了社团发现的精度。所挖掘的社团结构关系还表明,高校不同院系和学科形成的社团具有各自的特点。
[Abstract]:Social network has become one of the most important ways for modern people to exchange and exchange information online. As an example of domestic Renren, a large number of young people, especially students, take this as a platform to discuss topics of interest to each other. People are connected by many factors, such as learning relationship, working relationship, common interest and so on. The social network dominated by college students is more likely to be associated with the same schools, departments and institutes, thus showing community structure. In this paper, we use the CNM algorithm to verify the hypothesis based on Renren's real data. At the same time, we also use the structural knowledge of social network to improve the CNM algorithm and improve the accuracy of community discovery. The structure of the association also shows that the associations and disciplines formed by different colleges and universities have their own characteristics.
【作者单位】：北京大学软件与微电子学院;北京大学计算语言学教育部重点实验室;
【基金】：国家高技术研究发展计划(863)(2012AA011101) 国家自然科学基金(91024009) 国家社会科学基金(12&ZD227)
【分类号】：TP393.09
【正文快照】： 1引言随着互联网的蓬勃发展,社交网络在人际交往中发挥着越来越重要的作用。发现其中的社团关系,分析社团的属性,是一项非常有意义的工作。不少学者从不同角度对社团网络及其属性做了研究,出现了小世界特征[1]、幂律分布[2]和社团结构[3-6]等相关的成果。尤其是社团结构,它在

【共引文献】