基于层次覆盖的多层网络社团发现算法

A multilayer network community detection algorithm based on hierarchical cover

摘要:

如何检测多层网络的局部社团是近年来的热门问题之一。现有算法多针对于单层网络衡量指标的设计与改进，但节点往往处于多种复杂关系之中。为了精确的划分多层网络社团结构，一种基于层次覆盖的多层网络社团发现算法被提出。该算法结合RA相似度提取每层的内外连接的拓扑信息，并通过比较每层的拓扑信息关系来提取社团结构。实验结果表明，与CLECC和CLEDCC两种算法相比，提出的算法不仅降低了时间复杂度，而且在划分社团的准确度方面也有所提高，同时可以确定多层网络中无直接相连节点的划分关系。

Abstract:

How to detect the local community of multilayer network is one of the hot issues in recent years.Existing algorithms focus on the design and improvement of monolayer network，However, a node often exists in various relations.Therefore, how to find the community structure of multilayer network is a more meaningful research issue.A multilayer network community detection algorithm based on hierarchical cover is proposed .The algorithm combines the RA similarity to extract the topological information of the inner and outer connections of each layer, and extracts the community structure by comparing the topological information of each layer.The experimental part tests the effectiveness of the algorithm,The test results show that compared with CLECC and CLEDCC algorithms, the proposed algorithm not only reduces the time complexity, but also improves the accuracy of detecting the community.Meanwhile, we can determine the relationship between the nodes without direct connections in the multilayer network.