首页> 外文期刊>International Journal Information Theories and Applications >Detecting Communities from Networks: Comparison of Algorithms on Real and Synthetic Networks
【24h】

Detecting Communities from Networks: Comparison of Algorithms on Real and Synthetic Networks

机译:从网络检测社区:真实网络和合成网络上算法的比较

获取原文
           

摘要

Communities in real world complex networks correspond to hiddenstructures that are composed of nodes tightly connected among themselvesand weakly connected with other nodes in the network. There are variousapplications of automatic community detection in computer science, medicine,machine learning, sociology, etc. In this paper, we first present the existingcommunity detection algorithms and evaluation measures used in order toconsider the algorithms effectiveness. We then report a deep comparison of thealgorithms using both large scale real world complex networks and artificialnetworks generated from stochastic block model. We found that Louvainalgorithm is consistently the best across both the measures and the networks (8real world and many varied synthetic networks) we tested. Fast Greedy andLeading Eigenvector algorithms are also good alternatives. Moreover,compared to related work, our paper considers both more algorithms and morenetworks.
机译:现实世界中复杂网络中的社区对应于隐藏结构,这些隐藏结构由彼此之间紧密连接且与网络中其他节点之间弱连接的节点组成。自动社区检测在计算机科学,医学,机器学习,社会学等领域有广泛的应用。本文首先介绍了现有的社区检测算法和所采用的评估措施,以考虑算法的有效性。然后,我们报告使用大规模现实世界复杂网络和从随机块模型生成的人工网络对算法进行的深入比较。我们发现Louvainalgorithm在我们测试的度量和网络(8个真实世界和许多不同的合成网络)中始终是最好的。快速贪婪和领先特征向量算法也是不错的选择。此外,与相关工作相比,本文考虑了更多的算法和更多的网络。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号