Beyond the Academic Monoculture: A Unified Framework and Industrial Perspective for Attributed Graph Clustering

Yunhui Liu; Yue Liu; Yongchao Liu; Tao Zheng; Stan Z. Li; Xinwang Liu; Tieke He

cs.LGMar 21, 2026

Beyond the Academic Monoculture: A Unified Framework and Industrial Perspective for Attributed Graph Clustering

Authors: Yunhui Liu, Yue Liu, Yongchao Liu, Tao Zheng, Stan Z. Li, Xinwang Liu, Tieke He

Organizations: State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, China. · the School of Computing, National University of Singapore, Singapore. · Ant Group, Hangzhou, China. · Westlake University, Hangzhou, China. ·

Abstract

Attributed Graph Clustering (AGC) is a fundamental unsupervised task that partitions nodes into cohesive groups by jointly modeling structural topology and node attributes. While the advent of graph neural networks and self-supervised learning has catalyzed a proliferation of AGC methodologies, a widening chasm persists between academic benchmark performance and the stringent demands of real-world industrial deployment. To bridge this gap, this survey provides a comprehensive, industrially grounded review of AGC from three complementary perspectives. First, we introduce the Encode-Cluster-Optimize taxonomic framework, which decomposes the diverse algorithmic landscape into three orthogonal, composable modules: representation encoding, cluster projection, and optimization strategy. This unified paradigm enables principled architectural comparisons and inspires novel methodological combinations. Second, we critically examine prevailing evaluation protocols to expose the field's academic monoculture: a pervasive over-reliance on small, homophilous citation networks, the inadequacy of supervised-only metrics for an inherently unsupervised task, and the chronic neglect of computational scalability. In response, we advocate for a holistic evaluation standard that integrates supervised semantic alignment, unsupervised structural integrity, and rigorous efficiency profiling. Third, we explicitly confront the practical realities of industrial deployment. By analyzing operational constraints such as massive scale, severe heterophily, and tabular feature noise alongside extensive empirical evidence from our companion benchmark, we outline actionable engineering strategies. Furthermore, we chart a clear roadmap for future research, prioritizing heterophily-robust encoders, scalable joint optimization, and unsupervised model selection criteria to meet production-grade requirements.

Figures & tables

Fig. 2 : The Encode-Cluster-Optimize Taxonomic Framework for Attributed Graph Clustering. Left: Core ECO Framework Flow, showing the sequential data flow, key input/intermediate/output variables, and operational modules. Right: Detailed ECO Taxonomy Tree, providing a hierarchical classification of the methodologies across the three orthogonal modules.

Method	Venue	Encoder ( $\mathcal{E}$ )	Cluster ( $\mathcal{C}$ )	Optimize ( $\mathcal{O}$ )	Core Objective	Complexity
Non-Parametric & Decoupled Methods
AGC [ 35 ]	IJCAI’19	Simple Filtering	Spectral	Decoupled	Adaptive Smoothing Order Selection	$O(N^{2})$
SSGC [ 37 ]	ICLR’21	Multi-Filtering	K-Means	Decoupled	Diffusion Kernel Ensemble	$O(N+M)$
FGC [ 44 ]	SDM’22	Multi-Filtering	Spectral	Decoupled	Multi-hop Filter + Anchor Selection	$O(N^{2})$
GRACE [ 43 ]	TKDD’22	Simple Filtering	K-Means	Decoupled	Generalized Graph Laplacian Filter	$O(N+M)$
NAFS [ 45 ]	ICML’22	Multi-Filtering	K-Means	Decoupled	Multi-Strategy Smoothing Ensemble	$O(N+M)$

TABLE I : Comprehensive ECO Taxonomy of Representative Attributed Graph Clustering Methods. Complexity denotes the dominant time/space complexity, where

N=|\mathcal{V}|

and

M=|\mathcal{E}|

Fig. 3 : Illustration of the three coordination patterns in the ECO framework. Decoupled training applies

\mathcal{L}_{\text{rep}}

to pre-train the encoder, then applies the cluster projector post-hoc. Joint training simultaneously optimizes

\mathcal{L}_{\text{rep}}+\lambda\mathcal{L}_{\text{clust}}

. Hybrid coordination interleaves these phases via iterative self-training or confidence-gated feedback.

Scale	Dataset	Domain	#Nodes	#Edges	Avg. Deg.	#Feat.	Feat. Type	#Clus.	$\mathcal{H}_{e}$	$\mathcal{H}_{n}$
Tiny	Cora	Citation	2,708	10,556	3.9	1,433	Textual	7	0.81	0.83
Tiny	Photo	Co-purchase	7,650	238,162	31.1	745	Textual	8	0.83	0.84
Small	Physics	Co-author	34,493	495,924	14.4	8,415	Textual	5	0.93	0.92
	HM †	Co-purchase	46,563	21,461,990	460.9	120	Tabular	21	0.16	0.35
	Flickr	Social	89,250	899,756	10.1	500	Textual	7	0.32	0.32
Medium	ArXiv	Citation	169,343	1,166,243	6.9	128	Textual	40	0.65	0.64

TABLE II : Summary of Representative Datasets for Our Companion Benchmark [ 1 ] . The collection is categorized by scale, spanning five orders of magnitude from Tiny to Massive. The datasets cover a wide range of domains and feature types (Textual, Tabular).

\mathcal{H}_{e}

and

\mathcal{H}_{n}

denote edge homophily and node homophily, respectively. Datasets marked with

\dagger

are industrial graphs with complex tabular features and low homophily, representing the industrial evaluation frontier [ 23 ] .

Fig. 4 : Landscape of AGC benchmark datasets along two axes: edge homophily ( x -axis) and graph scale ( y -axis, log scale). Bubble color denotes feature modality and bubble size denotes cluster count

K

. The shaded region marks the Academic Monoculture Zone (

\mathcal{H}_{e}>0.70

, scale

{<}\,10^{5}

). Our benchmark [ 1 ] spans the full landscape across scale, homophily, feature modality, and cluster granularity.

Fig. 5 : An industrial perspective on attributed graph clustering , connecting real-world industrial deployment, common academic evaluation pitfalls, and proposed holistic solutions. It aims to bridge the gap between academic research and real-world industrial needs, divided into three interconnected parts: (I) Industrial AGC Scenarios and Deployment Challenges (Sec. VIII-A ), (II) Common Pitfalls in Academic AGC Evaluation (Sec. I , VII-A , VII-B ), (III) Proposed Holistic AGC Evaluation Standard and Key Industrial Deployment Strategies (Sec. VII-D and VIII-B - VIII-D )).

Category	Method	Reddit		MAG		Pokec		Products		WebTopic		Papers100M
Category	Method	Mem(GB)	Time(m)	Mem(GB)	Time(m)	Mem(GB)	Time(m)	Mem(GB)	Time(m)	Mem(GB)	Time(m)	Mem(GB)	Time(h)
Traditional	KMeans	-	0.27	-	6.59	-	3.09	-	5.87	-	3.98	-	0.05
Traditional	Node2Vec	0.70	22.67	1.90	77.58	4.04	236.53	15.05	64.94	11.80	75.92	7.20	12.38
Non- Parametric	SSGC	-	5.36	-	6.51	-	4.43	-	8.05	-	10.38	-	-
	SAGSC	-	5.99	-	23.35	-	4.72	-	5.29	-	2.77	-	-
	MS2CAG	-	0.16	-	15.10	-	13.68	-	1.38	-	16.81	-	-

TABLE IV : Efficiency profiling results for larger datasets on a single NVIDIA V100 GPU (32GB). Mem: Peak GPU Memory; Time: Total Training and Clustering Time. Results are drawn from our companion benchmark [ 1 ] .

Yunhui Liu is a Ph.D. student at Nanjing University. His research interests focus on graph machine learning and structured data analysis.

Table 9

Yue Liu is a Ph.D. student at the National University of Singapore. His research interests focus on self-supervised learning, with a particular emphasis on the safety and intelligence of foundation models, as well as graph learning.

Table 10

Yongchao Liu is a Staff Engineer at Ant Group, focusing on graph learning and agent systems. He previously held research positions at the Georgia Institute of Technology and the University of Mainz.

Table 11

Tao Zheng is a Professor at the Software Institute, Nanjing University. His research interests include knowledge graphs and large language models.

Table 12

Stan Z. Li is a Professor with Westlake University. He received the Ph.D. degree from Surrey University, UK. His research interests include pattern recognition, machine learning, and AI for Life Science and Drug Design.

Table 13

Xinwang Liu is a Professor with the College of Computer Science and Technology, National University of Defense Technology. He received the Ph.D. degree from the National University of Defense Technology.

Table 14

Tieke He is an Associate Professor at the Software Institute, Nanjing University. He received the Ph.D. degree from Nanjing University. His research interests include knowledge graphs and large language models.

Table 15

Graph Clustering

Graph clustering groups network nodes by connectivity, attributes, or shared roles to reveal communities and other mesoscale structure.

See all 78 papers

Feb 9, 2026

Bridging Academia and Industry: A Comprehensive Benchmark for Attributed Graph Clustering

Attributed Graph Clustering (AGC) is a fundamental unsupervised task that integrates structural topology and node attributes to uncover latent patterns in graph-structured data. Despite its significance in industrial applications such as fraud detection and user segmentation, a significant chasm persists between academic research and real-world deployment. Current evaluation protocols suffer from the small-scale, high-homophily citation datasets, non-scalable full-batch training paradigms, and a reliance on supervised metrics that fail to reflect performance in label-scarce environments. To bridge these gaps, we present PyAGC, a comprehensive, production-ready benchmark and library designed to stress-test AGC methods across diverse scales and structural properties. We unify existing methodologies into a modular Encode-Cluster-Optimize framework and, for the first time, provide memory-efficient, mini-batch implementations for a wide array of state-of-the-art AGC algorithms. Our benchmark curates 12 diverse datasets, ranging from 2.7K to 111M nodes, specifically incorporating industrial graphs with complex tabular features and low homophily. Furthermore, we advocate for a holistic evaluation protocol that mandates unsupervised structural metrics and efficiency profiling alongside traditional supervised metrics. Battle-tested in high-stakes industrial workflows at Ant Group, this benchmark offers the community a robust, reproducible, and scalable platform to advance AGC research towards realistic deployment. The code and resources are publicly available via GitHub (https://github.com/Cloudy1225/PyAGC), PyPI (https://pypi.org/project/pyagc), and Documentation (https://pyagc.readthedocs.io). We invite readers to our latest survey (https://arxiv.org/abs/2603.20829) for the unified framework and industrial perspective that complement this benchmark.

Yunhui Liu, Pengyu Qiu, Yu Xing +6

Nanjing University & Ant Group, Nanjing, China · Ant Group, Hangzhou, China · Nanjing University, Nanjing, China +2

Graph Clustering

Ask PDF

Oct 6, 2026

Towards One-for-All Foundation Model for Attributed Graph Clustering

Attributed graph clustering aims to discover node groups by jointly exploiting node attributes and graph topology, yet its unsupervised nature makes model selection and adaptation inherently difficult. Existing methods typically train and tune a separate model for each input graph, leading to costly and fragile pipelines that often fail to transfer across graphs with different feature spaces, structural patterns, and attribute-structure correlations. In this paper, we study a one-for-all alternative: can a single model be trained once and directly applied to diverse attributed graphs without graph-specific training, fine-tuning, or hyperparameter search? We propose OFAG, a foundation model for attributed graph clustering. Building upon Prior-data Fitted Networks, OFAG learns a reusable clustering inference strategy from synthetic attributed graphs generated under broad priors over latent clusters, node attributes, and graph structures. To handle incompatible feature spaces across graphs, OFAG adopts a dimension-agnostic signal-wise graph encoder that treats each feature channel as a graph signal and models its response to shared graph filters. The model is trained with a hyperspherical clustering objective, producing clustering-friendly node representations in a single forward pass at inference time. On ten datasets, one frozen OFAG model achieves the best mean performance and average rank across NMI, ACC, ARI, and F1, while completing all ten datasets in 12.43 minutes total---over 6* faster than the second-fastest baseline and nearly 28* faster than the second-best on clustering quality. Our code and pretrained checkpoint are available at https://github.com/Cloudy1225/OFAG, allowing practitioners to directly apply OFAG to their own attributed graph datasets without additional training or tuning.

Yunhui Liu, Xudong Jin, Kang Zhang +5

State Key Laboratory for Novel Software Technology Nanjing University Nanjing, China

Graph Representation Learning Graph Foundation Models Graph Clustering

Ask PDF

Jun 22, 2026

Bridge the Gaps: Heterogeneous Attributed Graph Clustering via Quaternion Representation Learning

Attributed graph clustering partitions nodes by jointly exploiting node attributes and graph topology. It remains challenging due to attribute heterogeneity and representation degradation during graph learning. Real-world datasets often contain heterogeneous attributes, i.e., numerical and categorical attributes, complicating unified representation learning. This challenge becomes more complex in attributed graphs, where constructing a clustering-friendly graph structure from attributes and topology remains difficult. Under deep graph architectures, repeated graph propagation causes node embeddings to become overly similar, leading to the over-smoothing (OS) effect. Meanwhile, graph representation learning amplifies topological influence, making discriminative attribute information harder to exploit for clustering, an effect we refer to as over-dominating (OD). To bridge these gaps, an end-to-end framework, Any-type attributed Graph REpresentation lEarning (AGREE), is proposed. It unifies attributed graphs and any-type attributed data through multi-level alignment and similarity-based graph construction. Quaternion-based graph convolution strengthens attribute interaction to alleviate OD, while shallow graph architectures help relieve OS. The learned embeddings are jointly optimized for graph reconstruction and clustering, without requiring a predefined number of clusters during training. Experiments on diverse benchmarks show that AGREE achieves strong overall performance in accuracy, robustness, and adaptability.

Xinxi Chen, Junyang Chen, Yiqun Zhang +2

School of Computer Science and Technology, Guangdong University of Technology, Guangzhou 510006, China

Graph Structure Learning Graph Neural Networks Graph Representation Learning Graph Clustering Oversmoothing

Ask PDF

Abstract

Graph Clustering

Explore similar work

Bridging Academia and Industry: A Comprehensive Benchmark for Attributed Graph Clustering

Towards One-for-All Foundation Model for Attributed Graph Clustering

Bridge the Gaps: Heterogeneous Attributed Graph Clustering via Quaternion Representation Learning