基于节点链接重要性的社区方法研究
Research on community methods based on node-links importance
朱叶 1杨帆 2章敏3
作者信息
- 1. 江西科技学院,江西 南昌 330098
- 2. 江西财经职业学院,江西 九江 332000
- 3. 九江职业技术学院,江西 九江 332099
- 折叠
摘要
关于社区结构较为常见的一种描述是关于强弱社区的阐述,一条边由 2 个节点构成,此 2 个节点对这条边有直接影响,是通过此2 个节点的邻接点数目直接影响.除此之外,其间接关联性是通过此2 个节点的公共邻居节点对网络其余部分的间接影响.文章从这 2 个方面考虑,得到一种新的社区发现算法.文章采用10 个社区结构清晰度不同的人工生成网络数据集,通过与几种社区发现算法进行对比实验.实验证明,从几种不同的评价指标角度而言,文章研究算法在社区结构清晰度不同的人工生成网络数据集上都有较好的结果.
Abstract
A common description of community structure is about strong and weak communities.An edge is composed of two nodes,which have a direct impact on the edge by the number of adjacent nodes.In addition,its indirect correlation is the indirect impact of the common neighboring nodes of these two nodes on the rest of the network.Based on these two considerations,a new community discovery algorithm is obtained.This article uses 10 manually generated network datasets with different clarity of community structures,and conducts comparative experiments with several community discovery algorithms.Experimental results have shown that the algorithm proposed in this paper achieves good results on artificially generated network datasets with different community structure clarity,from the perspective of several different evaluation indicators.
关键词
节点链接重要性/中心度/社区发现/复杂网络Key words
the importance of node-links/centrality/community discovery/complex network引用本文复制引用
基金项目
江西科技学院自然科学研究项目(23ZRYB05)
江西省教育厅科学技术研究项目(GJJ2202622)
出版年
2024
NETL
NSTL
万方数据