Awesome Attributed Graph Clustering (AGC) Papers
August 21, 2026 · View on GitHub
A curated list of papers on Attributed Graph Clustering (AGC). This list accompanies the survey paper Beyond the Academic Monoculture: A Unified Framework and Industrial Perspective for Attributed Graph Clustering and the benchmark paper Bridging Academia and Industry: A Comprehensive Benchmark for Attributed Graph Clustering.
Table of Contents
- Survey & Benchmark
- Research Papers
- Non-Parametric & Decoupled Methods
- Deep Decoupled Methods
- Deep Joint Methods
- Hybrid Coordination Methods
- Multi-View and Multimodal Graph Clustering
- Attributed Hypergraph Clustering
- Dynamic and Temporal Graph Clustering
- Attribute-Missing Graph Clustering
- Large-Scale and Scalable Methods
- LLM-Enhanced Graph Clustering
- Federated Graph Clustering
- Other Extensions
- Application Papers
- ECO Taxonomy Quick Reference
- Community Resources
- Citation
Survey & Benchmark
Research Papers
Papers are organized following the Encode-Cluster-Optimize framework introduced in the survey. Within each category, papers are sorted by year (descending).
Non-Parametric & Decoupled Methods
Encoders apply fixed spectral filtering operations without learnable weights; cluster projectors are applied post-hoc to frozen embeddings.
Deep Decoupled Methods
Parametric encoders are pre-trained with self-supervised representation objectives; cluster projectors are applied post-hoc to frozen embeddings.
Deep Joint Methods
Parametric encoders and cluster projectors are optimized simultaneously, allowing clustering objectives to directly shape the representation space.
Hybrid Coordination Methods
Methods that interleave decoupled pre-training and joint fine-tuning, using confidence-gated feedback, iterative self-training, or reinforcement-based coordination.
Multi-View and Multimodal Graph Clustering
Methods handling multiple graph views, attribute views, heterogeneous information networks, or multimodal attributed graphs.
Attributed Hypergraph Clustering
Methods for clustering on hypergraphs where hyperedges connect arbitrary subsets of nodes.
Dynamic and Temporal Graph Clustering
Methods for clustering on evolving or temporal graphs.
Attribute-Missing Graph Clustering
Methods handling partially or fully missing node attributes.
Large-Scale and Scalable Methods
Methods specifically designed for graphs with millions to billions of nodes, emphasizing linear-time complexity or mini-batch training.
LLM-Enhanced Graph Clustering
Methods leveraging large language models for text-attributed graph clustering.
| Year | Venue | Title | Code |
|---|---|---|---|
| 2026 | WWW | Node Role-Guided LLMs for Dynamic Graph Clustering | Code |
| 2025 | ACL Findings | MARK: Multi-agent Collaboration with Ranking Guidance for Text-attributed Graph Clustering | Code |
| 2024 | LoG | Large Language Model Guided Graph Clustering | — |
Federated Graph Clustering
Methods for privacy-preserving distributed graph clustering across multiple clients.
Other Extensions
Methods addressing specialized graph clustering settings including signed graphs, heterophilous graphs, and unknown cluster numbers.
Application Papers
Papers applying graph clustering techniques to downstream real-world tasks, organized by application domain.
Fraud Detection and Anomaly Detection
Recommendation Systems
| Year | Venue | Title | Code |
|---|---|---|---|
| 2025 | WWW | GraphHash: Graph Clustering Enables Parameter Efficiency in Recommender Systems | Code |
| 2024 | NeurIPS | End-to-end Learnable Clustering for Intent Learning in Recommendation | Code |
Graph Condensation and Distillation
| Year | Venue | Title | Code |
|---|---|---|---|
| 2026 | TKDE | DeepCGC: Unveiling the Deep Clustering Mechanism of Fast Graph Condensation | Code |
| 2025 | KDD | Simple yet Effective Graph Distillation via Clustering | Code |
Bioinformatics and Medical Science
Natural Language Processing
ECO Taxonomy Quick Reference
The following table summarizes representative methods under the Encode-Cluster-Optimize framework proposed in the survey. For each method, we list the encoder type, cluster projector, coordination pattern, and dominant complexity.
| Method | Venue | Encoder (E) | Cluster (C) | Coordinate (O) | Complexity |
|---|---|---|---|---|---|
| NP & Decoupled | |||||
| AGC | IJCAI'19 | Simple Filtering | Spectral | Decoupled | O(N²) |
| SSGC | ICLR'21 | Multi-Filtering | K-Means | Decoupled | O(N+M) |
| FGC | SDM'22 | Multi-Filtering | Spectral | Decoupled | O(N²) |
| GRACE | TKDD'22 | Simple Filtering | K-Means | Decoupled | O(N+M) |
| NAFS | ICML'22 | Multi-Filtering | K-Means | Decoupled | O(N+M) |
| SAGSC | AAAI'23 | Subspace-Oriented | Subspace | Decoupled | O(N+M) |
| IAGC | TKDE'23 | Simple Filtering | Spectral | Decoupled | O(N²) |
| SASE | CIKM'24 | Simple Filtering | Spectral | Decoupled | O(N+M) |
| S2CAG | KDD'25 | Subspace-Oriented | Subspace | Decoupled | O(N+M) |
| MS2CAG | KDD'25 | Subspace-Oriented | Subspace | Decoupled | O(N+M) |
| CMV-ND | ICML'25 | Multi-Filtering | K-Means | Decoupled | O(N+M) |
| Deep Decoupled | |||||
| GAE | NeurIPS-W'16 | GCN | K-Means | Decoupled | O(N²) |
| ARGA | IJCAI'18 | GCN | K-Means | Decoupled | O(N²) |
| DGI | ICLR'19 | GCN | K-Means | Decoupled | O(N+M) |
| MVGRL | ICML'20 | GCN | K-Means | Decoupled | O(N+M) |
| CCA-SSG | NeurIPS'21 | GCN | K-Means | Decoupled | O(N+M) |
| BGRL | ICLR'22 | GCN | K-Means | Decoupled | O(N+M) |
| DCRN | AAAI'22 | Mixed GNN | K-Means | Decoupled | O(N²) |
| S3GC | NeurIPS'22 | GCN | K-Means | Decoupled | O(N²) |
| DGCN | ICML'23 | Mixed GNN | K-Means | Decoupled | O(N+M) |
| NS4GC | TKDE'24 | GCN | K-Means | Decoupled | O(N²) |
| MAGI | KDD'24 | GCN/SAGE | K-Means | Decoupled | O(N²) |
| NeuCGC | TKDE'25 | GCN | K-Means | Decoupled | O(N²) |
| CoCo | ICLR'26 | Mixed GNN | K-Means | Decoupled | O(N+M) |
| Deep Joint | |||||
| DAEGC | IJCAI'19 | GAT | Prototype | Joint | O(N²) |
| SDCN | WWW'20 | GCN+AE | Prototype | Joint | O(N²) |
| MinCut | ICML'20 | GCN | Softmax | Joint | O(N+M) |
| RGAE | TKDE'22 | GCN | Prototype | Joint | O(N²) |
| DMoN | JMLR'23 | GCN | Softmax | Joint | O(N+M) |
| DinkNet | ICML'23 | GCN | Prototype | Joint | O(N+M) |
| DGCluster | AAAI'24 | GCN | Softmax | Joint | O(N+M) |
| Neuromap | NeurIPS'24 | GCN | Softmax | Joint | O(N+M) |
| LSEnet | ICML'24 | Lorentz GCN | Softmax (Tree) | Joint | O(N²) |
| GCSBM | LoG'25 | GCN | Softmax | Joint | O(N+M) |
| DeSE | KDD'25 | GCN | Softmax | Joint | O(N²) |
| FVD | TPAMI'25 | GCN+VAE | Prototype | Joint | O(N²) |
| ASIL | TPAMI'26 | Lorentz GCN | Softmax (Tree) | Joint | O(N²) |
| Hybrid Coord. | |||||
| CLEAR | TIST'23 | GCN | K-Means | Hybrid | O(N+M) |
| HoLe | CIKM'23 | GCN | K-Means | Hybrid | O(N²) |
| HSAN | AAAI'23 | GCN | K-Means | Hybrid | O(N²) |
| CCGC | AAAI'23 | GCN | K-Means | Hybrid | O(N²) |
| RGC | MM'23 | GCN | K-Means | Hybrid | O(N²) |
| CARL-G | KDD'23 | GCN | Softmax | Hybrid | O(N+M) |
| NeuroCUT | KDD'24 | GNN | Softmax | Hybrid | O(N+M) |
| FastDGC | TKDD'24 | GCN | K-Means | Hybrid | O(N²) |
| DGAC | WWW'25 | Mixed GNN | K-Means | Hybrid | O(N²) |
| RAGC | NeurIPS'25 | Mixed GNN | K-Means | Hybrid | O(N²) |
Community Resources
A collection of actively maintained repositories, benchmarks, and reading lists related to attributed graph clustering and deep graph clustering.
| Resource | Description |
|---|---|
| PyAGC Reading List | Curated paper list accompanying the PyAGC benchmark, covering AGC methods, datasets, and applications |
| PyAGC | Production-ready benchmark library for attributed graph clustering with standardized implementations, mini-batch support, and industrial-scale evaluation |
| Awesome Deep Graph Clustering | Comprehensive collection of deep graph clustering papers, codes, and datasets, accompanying this survey |
| PyDGC | Open-source Python library for deep graph clustering, accompanying the DGCBench benchmark with unified training and evaluation paradigms |
| BenchTGC | Benchmark repository for temporal graph clustering with curated datasets and standardized evaluation frameworks |
Citation
If you find this list useful, please consider citing our survey and benchmark papers:
@article{liu2026beyond,
title={Beyond the Academic Monoculture: A Unified Framework and Industrial Perspective for Attributed Graph Clustering},
author={Yunhui Liu and Yue Liu and Yongchao Liu and Tao Zheng and Stan Z. Li and Xinwang Liu and Tieke He},
year={2026},
eprint={2603.20829},
archivePrefix={arXiv},
primaryClass={cs.LG}
}
@article{liu2026bridging,
title={Bridging Academia and Industry: A Comprehensive Benchmark for Attributed Graph Clustering},
author={Yunhui Liu and Pengyu Qiu and Yu Xing and Yongchao Liu and Peng Du and Chuntao Hong and Jiajun Zheng and Tao Zheng and Tieke He},
year={2026},
eprint={2602.08519},
archivePrefix={arXiv},
primaryClass={cs.LG}
}