摘要
针对Louvain算法的分辨率存在极限的问题,提出一种基于多分辨率模块度(multi-resolution modularity,MRQ)的Louvain社区检测算法,在Louvain算法中加入一个可变分辨率模块MRQ,有效解决了Louvain算法分辨率存在极限的问题。在此基础上,针对Louvain算法会得出某些不满足社区条件的划分结果,增加社区划分的约束条件,提高了社区划分的精度。最后,在真实数据集和计算机生成网络上对该算法进行实验验证。实验结果表明,具有MRQ的约束Louvain算法在模块度、覆盖度和调制度指标上都优于Louvain、标签传播(LPA)、Kernighan-Lin(KL)和贪心模块度等主流社区检测算法。
Aiming at the problem that the resolution of the Louvain algorithm is limited,this paper proposed Louvain community detection algorithm based on multi-resolution modularity(MRQ).It added a variable resolution module MRQ to the Louvain algorithm to effectively solve the problem of the resolution limit of the Louvain algorithm.On this basis,to solve the problem that the Louvain algorithm could obtain division results that did not meet the community conditions,the constraints of community division were added,and the accuracy of the community division was improved.Finally,it experimentally verified the algorithm on real datasets and computer-generated networks.The experimental results show that the constrained Louvain algorithm with MRQ outperforms mainstream community detection algorithms such as the Louvain algorithm,label propagation algorithm(LPA),Kernighan-Lin(KL)algorithm and greedy modularity algorithm in terms of modularity,coverage and modulation index.
作者
张震
靳金帅
陈可鑫
田鸿朋
Zhang Zhen;Jin Jinshuai;Chen Kexin;Tian Hongpeng(School of Electrical&Information Engineering,Zhengzhou University,Zhengzhou 450001,China;Zhuhai Unitech Power Technology Co.,Ltd.,Zhuhai Guangdong 519000,China)
出处
《计算机应用研究》
北大核心
2025年第8期2335-2340,共6页
Application Research of Computers
基金
河南省重大公益专项资助项目(201300311200)
河南省重点研发专项资助项目(231111211600)
教育部产学合作协同育人项目(220802377143931)。
