GATFELPA integrates graph attention networks and enhanced label propagation for robust community detection
Abstract Community detection in graph networks is a fundamental yet challenging task due to limitations in existing methods. Traditional Graph Convolutional Network (GCN)–based models often suffer from over-smoothing, which results in indistinguishable node representations after excessive aggregatio...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-01-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-024-84962-4 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832571827665240064 |
|---|---|
| author | Feiyi Tang Junxian Li Xi Liu Chao Chang Luyao Teng |
| author_facet | Feiyi Tang Junxian Li Xi Liu Chao Chang Luyao Teng |
| author_sort | Feiyi Tang |
| collection | DOAJ |
| description | Abstract Community detection in graph networks is a fundamental yet challenging task due to limitations in existing methods. Traditional Graph Convolutional Network (GCN)–based models often suffer from over-smoothing, which results in indistinguishable node representations after excessive aggregation. Additionally, many models face computational inefficiency, restricting their scalability on large-scale networks. To address these challenges, we propose GATFELPA, a hybrid community detection model that combines a Graph Attention Network (GAT) with an enhanced label propagation algorithm (DPCELPA). GATFELPA employs an adaptive strategy to dynamically determine the optimal number of aggregation layers, effectively mitigating over-smoothing by balancing intra-community compactness and inter-community separability. A novel similarity preservation module further enhances the model’s ability to differentiate communities by retaining local and global dissimilarities in heterogeneous networks. Comparative experiments on four real-world datasets, including large-scale networks like ogbn-arxiv, demonstrate that GATFELPA achieves superior performance across most metrics, particularly excelling in accuracy, robustness, and scalability. These results highlight GATFELPA as a promising approach for addressing complex community detection tasks. |
| format | Article |
| id | doaj-art-d76c2f1a9e97461283c59dae78fb808f |
| institution | Kabale University |
| issn | 2045-2322 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-d76c2f1a9e97461283c59dae78fb808f2025-02-02T12:16:48ZengNature PortfolioScientific Reports2045-23222025-01-0115111710.1038/s41598-024-84962-4GATFELPA integrates graph attention networks and enhanced label propagation for robust community detectionFeiyi Tang0Junxian Li1Xi Liu2Chao Chang3Luyao Teng4School of Information Engineering, Guangzhou Panyu PolytechnicSchool of Information Engineering, Guangzhou Panyu PolytechnicSchool of Information Engineering, Guangzhou Panyu PolytechnicSchool of Information Engineering, Guangzhou Panyu PolytechnicSchool of Information Engineering, Guangzhou Panyu PolytechnicAbstract Community detection in graph networks is a fundamental yet challenging task due to limitations in existing methods. Traditional Graph Convolutional Network (GCN)–based models often suffer from over-smoothing, which results in indistinguishable node representations after excessive aggregation. Additionally, many models face computational inefficiency, restricting their scalability on large-scale networks. To address these challenges, we propose GATFELPA, a hybrid community detection model that combines a Graph Attention Network (GAT) with an enhanced label propagation algorithm (DPCELPA). GATFELPA employs an adaptive strategy to dynamically determine the optimal number of aggregation layers, effectively mitigating over-smoothing by balancing intra-community compactness and inter-community separability. A novel similarity preservation module further enhances the model’s ability to differentiate communities by retaining local and global dissimilarities in heterogeneous networks. Comparative experiments on four real-world datasets, including large-scale networks like ogbn-arxiv, demonstrate that GATFELPA achieves superior performance across most metrics, particularly excelling in accuracy, robustness, and scalability. These results highlight GATFELPA as a promising approach for addressing complex community detection tasks.https://doi.org/10.1038/s41598-024-84962-4Community detectionGraph attention network (GAT)Density peak clusteringLabel propagation algorithm (LPA) |
| spellingShingle | Feiyi Tang Junxian Li Xi Liu Chao Chang Luyao Teng GATFELPA integrates graph attention networks and enhanced label propagation for robust community detection Scientific Reports Community detection Graph attention network (GAT) Density peak clustering Label propagation algorithm (LPA) |
| title | GATFELPA integrates graph attention networks and enhanced label propagation for robust community detection |
| title_full | GATFELPA integrates graph attention networks and enhanced label propagation for robust community detection |
| title_fullStr | GATFELPA integrates graph attention networks and enhanced label propagation for robust community detection |
| title_full_unstemmed | GATFELPA integrates graph attention networks and enhanced label propagation for robust community detection |
| title_short | GATFELPA integrates graph attention networks and enhanced label propagation for robust community detection |
| title_sort | gatfelpa integrates graph attention networks and enhanced label propagation for robust community detection |
| topic | Community detection Graph attention network (GAT) Density peak clustering Label propagation algorithm (LPA) |
| url | https://doi.org/10.1038/s41598-024-84962-4 |
| work_keys_str_mv | AT feiyitang gatfelpaintegratesgraphattentionnetworksandenhancedlabelpropagationforrobustcommunitydetection AT junxianli gatfelpaintegratesgraphattentionnetworksandenhancedlabelpropagationforrobustcommunitydetection AT xiliu gatfelpaintegratesgraphattentionnetworksandenhancedlabelpropagationforrobustcommunitydetection AT chaochang gatfelpaintegratesgraphattentionnetworksandenhancedlabelpropagationforrobustcommunitydetection AT luyaoteng gatfelpaintegratesgraphattentionnetworksandenhancedlabelpropagationforrobustcommunitydetection |