Community Detection by Fuzzy Relations
Community Detection by Fuzzy Relations
复制标题
DOI：
10.1109/tetc.2017.2751101
复制
发表时间：
2020-04-01
影响因子：
5.9
通讯作者：
Zhang, Daofu
中科院分区：
文献类型：
--
作者：
Luo, Wenjian;Yan, Zhenglong;Zhang, Daofu
网络数据对知识的需求不断增长，在许多任务中构成了重大挑战。从网络中发现社区结构是网络分析中面临的经典问题之一。在本文中，我们从模糊关系的组成的角度研究了网络结构，并提出了一种基于模糊关系的新算法，即CDFR（通过模糊关系的社区检测）提出了非重叠的社区检测。 CDFR的关键思想是找到每个节点的NGC节点（最接近中心性），并计算它们之间的模糊关系。然后，一个节点所属的社区取决于其NGC节点。此外，将构建决策图以指导社区检测。对人工和现实世界网络的实验结果验证了我们CDFR算法的有效性和优势。
The increasing demand for knowledge from network data poses significant challenges in many tasks. Discovering community structure from a network is one of the classic and significant problems faced in network analysis. In this paper, we study the network structure from the perspective of the composition of fuzzy relations, and a novel algorithm based on fuzzy relations, i.e., CDFR (Community Detection by Fuzzy Relations), is proposed for non-overlapping community detection. The key idea of CDFR is to find the NGC node (Nearest node with Greater Centrality) for each node and compute the fuzzy relation between them. Then, the community to which a node belongs depends on its NGC node. In addition, the decision graph will be constructed to guide community detection. Experimental results on artificial and real-world networks verify the effectiveness and superiority of our CDFR algorithm.