基于形式概念分析的博客社区发现
发布时间：2018-08-20 19:13
【摘要】：针对拖网算法存在的发现Web社区数量过多、社区间页面重复率较高以及严格的社区定义形成孤立社区等问题,提出一种基于形式概念分析(FCA)的博客社区发现算法。根据博客网络之间的链接关系构造概念格,通过格的代数消解对原始概念格进行等价划分,度量每个划分中概念间外延和内涵的结构相似性进而合并社区核心形成社区。实验结果表明:测试数据集中社区核心的网络密度大于40%的占全部的83.420%,合并社区的网络直径为3,且社区内容丰富程度得到提高。所提算法可以有效地运用于博客、微博等社交网络的社区发现,具有显著的应用价值和现实意义。
[Abstract]:In order to solve the problems such as excessive number of Web communities, high page repetition rate among communities and strict community definition to form isolated communities, a blog community discovery algorithm based on formal concept analysis (FCA) is proposed. The concept lattice is constructed according to the link relation between the blog network, and the original concept lattice is divided by algebraic deconstruction of the lattice, and the structural similarity of the extension and connotation among the concepts in each partition is measured, and then the community core is merged to form the community. The experimental results show that the network density of the community core in the test data set is more than 40%, the network diameter of the merged community is 3, and the richness of the community content is improved. The proposed algorithm can be effectively applied to social networks such as blog, Weibo and so on, which has significant application value and practical significance.
【作者单位】： 苏州大学计算机科学与技术学院;
【基金】：国家自然科学基金资助项目(61070122)
【分类号】：TP393.092
[Abstract]:In order to solve the problems such as excessive number of Web communities, high page repetition rate among communities and strict community definition to form isolated communities, a blog community discovery algorithm based on formal concept analysis (FCA) is proposed. The concept lattice is constructed according to the link relation between the blog network, and the original concept lattice is divided by algebraic deconstruction of the lattice, and the structural similarity of the extension and connotation among the concepts in each partition is measured, and then the community core is merged to form the community. The experimental results show that the network density of the community core in the test data set is more than 40%, the network diameter of the merged community is 3, and the richness of the community content is improved. The proposed algorithm can be effectively applied to social networks such as blog, Weibo and so on, which has significant application value and practical significance.
【作者单位】： 苏州大学计算机科学与技术学院;
【基金】：国家自然科学基金资助项目(61070122)
【分类号】：TP393.092
【参考文献】
相关期刊论文 前4条
1 杨楠,弓丹志,李_，
本文编号：2194710
本文链接：https://www.wllwen.com/guanlilunwen/ydhl/2194710.html
