Toward network clustering using graph-boosting and graph-factorization

Yongqiang Gao and Amin Rezaeipanah

Chaos, Solitons & Fractals, 2026, vol. 202, issue P1

Abstract: Network clustering using contrastive learning is a cutting-edge technique that leverages the principles of self-supervised learning to identify meaningful clusters within a network. Contrastive learning is used to learn meaningful representations of nodes in the network. These representations capture important characteristics of the nodes that can be used for clustering. However, there are three problems for this learning method during network clustering: Firstly, the efficiency of positive samples depends on the carefully constructed data augmentation. Secondly, the relationships between distant nodes are neglected due to the sparseness of the network. Thirdly, the model's pre-training and fine-tuning stages must handle large-scale datasets. To overcome these challenges, this study develops an efficient Network Clustering method using Graph-Boosting and Graph-Factorization (NC-GBGF). Specifically, the graph-boosting module includes auxiliary, enhance, and refine graphs to apply distant relationships between nodes and improve the representation of global relationships in the clustering process. Meanwhile, the graph-factorization module includes a singular value decomposition approach for extracting primary patterns and distinguishable features from large-scale data. Also, a pseudo-Siamese neural network based on dual-guidance supervisor is configured in NC-GBGF for model training. Ultimately, the final clustering is produced using a subspace clustering framework. Comprehensive experiments on several real-world datasets confirm the superiority of NC-GBGF compared to state-of-the-art algorithms. Our method shows significant advantages in deep graph clustering, because it simplifies network structure, augments data, manages sparse and large-scale data, and avoids essential information loss.

Keywords: Contrastive learning; Network clustering; Graph-boosting; Graph-factorization; Subspace clustering; Dual-guidance supervisor (search for similar items in EconPapers)
Date: 2026
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S0960077925015383
Full text for ScienceDirect subscribers only

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:eee:chsofr:v:202:y:2026:i:p1:s0960077925015383

DOI: 10.1016/j.chaos.2025.117525

Access Statistics for this article

Chaos, Solitons & Fractals is currently edited by Stefano Boccaletti and Stelios Bekiros

More articles in Chaos, Solitons & Fractals from Elsevier
Bibliographic data for series maintained by Thayer, Thomas R. ().