.png){kind=icon}
返回
Compactness Preserving Community Computation Via a Network Generative Process
.png){kind=icon}
PRE
AI
DOI：10.1109/TETCI.2021.3110086.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
The research on network community structure recently has some emerging aspects, such as community hiding for privacy protection in social networks, information transmission between different functional structures in brain networks and community structure's stability in ecological networks. However, all these attractive topics require us to drill down to the understanding of community structure, among which the key is how to define a reliable community and carry out an effective community computation. The goal of community computation is to find a community structure efficiently, in which nodes in a common community are densely connected, while nodes distributed in different communities are sparsely connected. Here we first define a new metric called compactness to measure the significance of a network's community structure. For a target community, compactness is able to differentiate the impact of other nearby communities to the target community on network's community significance, meanwhile the size of these communities is also taken into consideration. Then, we extend compactness and develop a fuzzy compactness via a network generation process to measure a fuzzy community structure's significance where the impact of other nearby communities to the target community can be differentiated in a differentiable manner. Further, we propose FCOCD-an effective approach with fuzzy compactness optimization for overlapping community detection in networks. Comprehensive experiments were conducted on several real-world and synthetic networks (including a case study on a bottlenose dolphin ecological network), and the results show that FCOCD can yield good performance in terms of detection accuracy as well as scalability.
Keyword:
Measurement
Stability analysis
Task analysis
Computational intelligence
Toy manufacturing industry
Tools
Social networking (online)
Community computation
compactness
nearby community differentiation
network generative process
期刊
I
IF:
6.5
论文数:
1.4K
被引数:
4.5K
机构
暂无机构信息
引用论文
.png){kind=icon}
分享收藏Graph K-means Based on Leader Identification, Dynamic Game, and Opinion Dynamics基于领导者识别、动态博弈和意见动态的k-means图
PREAI
分享收藏分享收藏
分享收藏Tinnitus Retraining Therapy (TRT) as a Method for Treatment of Tinnitus and Hyperacusis Patients耳鸣再训练疗法 (TRT) 作为治疗耳鸣和高亢患者的方法
PREAI
分享收藏
查看更多内容