- Physical Review Journals
All JournalsPhysics Magazine

Physical Review Research

Reuse & Permissions

It is not necessary to obtain permission to reuse this article or its components as it is available under the terms of the Creative Commons Attribution 4.0 International license. This license permits unrestricted use, distribution, and reproduction in any medium, provided attribution to the author(s) and the published article's title, journal citation, and DOI are maintained. Please note that some figures may have been included with permission from other third parties. It is your responsibility to obtain the proper permission from the rights holder directly for these figures.

Export citation

Export citation

Choose format for download:

Download Citation
  • Open Access

Community detection robustness of graph neural networks

Jaidev Goel1,*, Pablo Moriano2,†,‡, Ramakrishnan Kannan2, and Yulia R. Gel1

PDF

Phys. Rev. Research 8, 023123 – Published 5 May, 2026

DOI: https://doi.org/10.1103/v6nq-hcm8

Abstract

Graph neural networks (GNNs) are increasingly widely used for community detection in attributed networks. They combine structural topology with node attributes through message passing and pooling. However, their robustness or lack thereof with respect to different perturbations and targeted attacks in conjunction with community detection tasks is not well understood. To shed light on latent mechanisms behind GNN sensitivity on community detection tasks, we conduct a systematic computational evaluation of six widely adopted GNN architectures graph convolutional network, graph attention network, graph sample and aggregate (GraphSAGE), differentiable pooling (DiffPool), minimum cut pooling (MinCUT), and deep modularity networks (DMoN). The analysis covers three perturbation categories: node attribute manipulations, edge topology distortions, and adversarial attacks. We use element-centric similarity as the evaluation metric on synthetic benchmarks and real-world citation networks. Our findings indicate that supervised GNNs tend to achieve higher baseline accuracy, while unsupervised methods, particularly DMoN, maintain stronger resilience under targeted and adversarial perturbations. Furthermore, robustness appears to be strongly influenced by community strength, with well-defined communities reducing performance loss. Across all models, node attribute perturbations associated with targeted edge deletions and shifts in attribute distributions tend to cause the largest degradation in community recovery. These findings highlight important trade-offs between accuracy and robustness in GNN-based community detection and offer insights into selecting architectures resilient to noise and adversarial attacks.

View figure in article

Physics Subject Headings (PhySH)

Article Text

References (61)

  1. S. Fortunato, Community detection in graphs, Phys. Rep. 486, 75 (2010).
  2. M. E. Newman, Modularity and community structure in networks, Proc. Natl. Acad. Sci. USA 103, 8577 (2006).
  3. M. Girvan and M. E. Newman, Community structure in social and biological networks, Proc. Natl. Acad. Sci. USA 99, 7821 (2002).
  4. S. Wandelt, X. Shi, and X. Sun, Estimation and improvement of transportation network robustness by exploiting communities, Reliab. Eng. Syst. Saf. 206, 107307 (2021).
  5. A. Lancichinetti and S. Fortunato, Community detection algorithms: A comparative analysis, Phys. Rev. E 80, 056117 (2009).
  6. P. Moriano, J. Finke, and Y.-Y. Ahn, Community-based event detection in temporal networks, Sci. Rep. 9, 4358 (2019).
  7. S. Aref and M. Mostajabdaveh, Analyzing modularity maximization in approximation, heuristic, and graph neural network algorithms for community detection, J. Comput. Sci. 78, 102283 (2024).
  8. T. Schuurman and E. Bruner, Modularity and community detection in human brain morphology, Anat. Rec. 307, 345 (2024).
  9. F. R. Khawaja, Z. Zhang, Y. Memon, and A. Ullah, Exploring community detection methods and their diverse applications in complex networks: A comprehensive review, Soc. Network Anal. Min. 14, 115 (2024).
  10. F. Scarselli, M. Gori, A. C. Tsoi, M. Hagenbuchner, and G. Monfardini, The graph neural network model, IEEE Trans. Neural Netw. 20, 61 (2008).
  11. Z. Wu, S. Pan, F. Chen, G. Long, C. Zhang, and P. S. Yu, A comprehensive survey on graph neural networks, IEEE Trans. Neural Netw. Learn. Syst. 32, 4 (2020).
  12. K. Xu, W. Hu, J. Leskovec, and S. Jegelka, How powerful are graph neural networks? in International Conference on Learning Representations (ICLR, 2019).
  13. P. Veličković, Everything is connected: Graph neural networks, Curr. Opin. Struct. Biol. 79, 102538 (2023).
  14. T. N. Kipf and M. Welling, Semi-supervised classification with graph convolutional networks, in International Conference on Learning Representations (ICLR, 2017).
  15. P. Veličković, G. Cucurull, A. Casanova, A. Romero, P. Lio, and Y. Bengio, Graph attention networks, in International Conference on Learning Representations (ICLR, 2018).
  16. W. Hamilton, Z. Ying, and J. Leskovec, Inductive representation learning on large graphs, in Advances in Neural Information Processing Systems (Curran Associates Inc., Red Hook, NY, 2017), Vol. 30, pp. 1024–1034.
  17. X. Su, S. Xue, F. Liu, J. Wu, J. Yang, C. Zhou, W. Hu, C. Paris, S. Nepal, D. Jin, et al., A comprehensive survey on community detection with deep learning, IEEE Trans. Neural Netw. Learn. Syst. 35, 4682 (2024).
  18. Y. Ren, J. Pu, Z. Yang, J. Xu, G. Li, X. Pu, S. Y. Philip, and L. He, Deep clustering: A comprehensive survey, IEEE Trans. Neural Netw. Learn. Syst. 36, 5858 (2025).
  19. Y. Liu, J. Xia, S. Zhou, X. Yang, K. Liang, C. Fan, Y. Zhuang, S. Z. Li, X. Liu, and K. He, A survey of deep graph clustering: Taxonomy, challenge, application, and open resource, arXiv:2211.12875.
  20. Z. Ying, J. You, C. Morris, X. Ren, W. Hamilton, and J. Leskovec, Hierarchical graph representation learning with differentiable pooling, in Advances in Neural Information Processing Systems (Curran Associates Inc., Red Hook, NY, 2018), Vol. 31, pp. 4800–4810.
  21. F. M. Bianchi, D. Grattarola, and C. Alippi, Spectral clustering with graph neural networks for graph pooling, in International Conference on Machine Learning (PMLR, 2020), pp. 874–883.
  22. A. Tsitsulin, J. Palowitch, B. Perozzi, and E. Müller, Graph clustering with graph neural networks, J. Mach. Learn. Res. 24, 1 (2023).
  23. D. Zügner, A. Akbarnejad, and S. Günnemann, Adversarial attacks on neural networks for graph data, in Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining (Association for Computing Machinery, New York, 2018), pp. 2847–2856.
  24. D. Zügner and S. Günnemann, Adversarial attacks on graph neural networks via meta learning, in International Conference on Learning Representations (ICLR, 2019).
  25. B. Wang, J. Jia, X. Cao, and N. Z. Gong, Certified robustness of graph neural networks against adversarial structural perturbation, in Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining (Association for Computing Machinery, New York, 2021), pp. 1645–1653.
  26. S. Günnemann, Graph neural networks: Adversarial robustness, in Graph Neural Networks: Foundations, Frontiers, and Applications (Springer, 2022), pp. 149–176.
  27. B. Zhang, X. Guo, Z. Tu, and J. Zhang, Graph alternate learning for robust graph neural networks in node classification, Neural Comput. Appl. 34, 8723 (2022).
  28. H. Zhang, B. Wu, X. Yuan, S. Pan, H. Tong, and J. Pei, Trustworthy graph neural networks: Aspects, methods, and trends, Proc. IEEE 112, 97 (2024).
  29. L. Sun, Y. Dou, C. Yang, K. Zhang, J. Wang, S. Y. Philip, L. He, and B. Li, Adversarial attack and defense on graph data: A survey, IEEE. Trans. Knowl. Data. Eng. 35, 7693 (2023).
  30. M. Tian and P. Moriano, Robustness of community structure under edge addition, Phys. Rev. E 108, 054302 (2023).
  31. Z.-F. Wei, P. Moriano, and R. Kannan, Robustness of graph embedding methods for community detection, arXiv:2405.00636.
  32. Y. Shang, Unveiling robustness and heterogeneity through percolation triggered by random-link breakdown, Phys. Rev. E 90, 032820 (2014).
  33. Y. Shang, Subgraph robustness of complex networks under attacks, IEEE Trans. Syst. Man Cybern.: Syst. 49, 821 (2017).
  34. Y. Shang, Attack robustness and stability of generalized k-cores, New J. Phys. 21, 093013 (2019).
  35. A. Lancichinetti, S. Fortunato, and F. Radicchi, Benchmark graphs for testing community detection algorithms, Phys. Rev. E 78, 046110 (2008).
  36. B. Karrer and M. E. J. Newman, Stochastic blockmodels and community structure in networks, Phys. Rev. E 83, 016107 (2011).
  37. A. K. Dey, Y. Tian, and Y. R. Gel, Community detection in complex networks: From statistical foundations to data science applications, Wiley Interdiscip. Rev. Comput. Stat. 14, e1566 (2022).
  38. Y. Zhao, E. Levina, and J. Zhu, Consistency of community detection in networks under degree-corrected stochastic block models, Ann. Stat. 40, 2266 (2012).
  39. A. Tsitsulin, B. Rozemberczki, J. Palowitch, and B. Perozzi, Synthetic graph generation to benchmark graph learning, arXiv:2204.01376.
  40. A. J. Gates, I. B. Wood, W. P. Hetrick, and Y.-Y. Ahn, Element-centric clustering comparison unifies overlaps and hierarchy, Sci. Rep. 9, 8574 (2019).
  41. J. Goel, Robust graph neural networks, GitHub (2025), https://github.com/jdg-stat/robustgnn.
  42. A. Lancichinetti and S. Fortunato, Benchmarks for testing community detection algorithms on directed and weighted graphs with overlapping communities, Phys. Rev. E 80, 016118 (2009).
  43. A. Hagberg, P. J. Swart, and D. A. Schult, Exploring network structure, dynamics, and function using NetworkX, Technical Report, 2008, Los Alamos National Laboratory (LANL), Los Alamos, NM.
  44. P. W. Holland, K. B. Laskey, and S. Leinhardt, Stochastic blockmodels: First steps, Soc. Netw. 5, 109 (1983).
  45. Z. Yang, W. Cohen, and R. Salakhudinov, Revisiting semi-supervised learning with graph embeddings, in International Conference on Machine Learning (PMLR, 2016), pp. 40–48.
  46. Y. Xu, M. Lanier, A. Sarkar, and Y. Vorobeychik, Attacks on node attributes in graph neural networks, arXiv:2402.12426.
  47. S. Souravlas, S. Anastasiadou, and S. Katsavounis, A survey on the recent advances of deep community detection, Appl. Sci. 11, 7179 (2021).
  48. S. Sobolevsky and A. Belyi, Graph neural network inspired algorithm for unsupervised network community detection, Appl. Network Sci. 7, 63 (2022).
  49. W. Leeney and R. McConville, Uncertainty in GNN learning evaluations: A comparison between measures for quantifying randomness in GNN community detection, Entropy 26, 78 (2024).
  50. V. D. Kumar, S. Sreesankar, K. Dinesan, P. B. Nair, and L. Deepthi, Analyzing GNN models for community detection using graph embeddings: A comparative study, in 2024 15th International Conference on Computing Communication and Networking Technologies (ICCCNT) (IEEE, 2024), pp. 1–8.
  51. C. Merkwirth and T. Lengauer, Automatic generation of complementary descriptors with molecular graph networks, J. Chem. Inf. Model. 45, 1159 (2005).
  52. G. Budel, M. Kitsak, R. Aldecoa, K. Zuev, and D. Krioukov, Random hyperbolic graphs in d + 1 dimensions, Phys. Rev. E 109, 054131 (2024).
  53. B. Désy, P. Desrosiers, and A. Allard, Dimension matters when modeling network communities in hyperbolic spaces, PNAS Nexus 2, pgad136 (2023).
  54. M. Yang, M. Zhou, Z. Li, J. Liu, L. Pan, H. Xiong, and I. King, Hyperbolic graph neural networks: A review of methods and applications, arXiv:2202.13852.
  55. N. A. Arafat, D. Basu, Y. Gel, and Y. Chen, When witnesses defend: A witness graph topological layer for adversarial graph learning, Proc. AAAI Conf. Artif. Intell. 39, 15408 (2025).
  56. Y. Chen, B. Coskunuzer, and Y. Gel, Topological relational learning on graphs, in Advances in Neural Information Processing Systems (NeurIPS, 2021), Vol. 34, pp. 27029–27042.
  57. Y. Chen and Y. R. Gel, Topological pooling on graphs, Proc. AAAI Conf. Artif. Intell. 37, 7096 (2023).
  58. Y. Huang, Y. Zeng, Q. Wu, and L. Lü, Higher-order graph convolutional network with flower-petals laplacians on simplicial complexes, Proc. AAAI Conf. Artif. Intell. 38, 12653 (2024).
  59. E. Isufi, G. Leus, B. Beferull-Lozano, S. Barbarossa, and P. Di Lorenzo, Topological signal processing and learning: Recent advances and future challenges, Signal Process. 233, 109930 (2025).
  60. Z. Yan, Q. Zhao, Z. Ye, T. Ma, L. Gao, Z. Tang, Y. Wang, and C. Chen, Enhancing graph representation learning with localized topological features, J. Mach. Learn. Res. 26, 1 (2025).
  61. http://energy.gov/downloads/doe-public-access-plan.

Outline

Information

Phys. Rev. Research 8, 023123– Published 5 May, 2026
  • Received 1 October 2025
  • Accepted 17 March 2026
Crossmark - Check for updates button
Creative Commons Logo

Published by the American Physical Society under the terms of the Creative Commons Attribution 4.0 International license. Further distribution of this work must maintain attribution to the author(s) and the published article's title, journal citation, and DOI.

Published by the American Physical Society

Sign In to Your Journals Account

Filter

Filter

Article Lookup

Enter a citation