knowledge-based Fuzzy Defensive Classi cation Model for Robust MPGNNs

Bastami, Esmaeil; Soltanizadeh, Hadi; Rahmanimanesh, Mohammad; Keshavarzi, Parviz; Asadpour, Masoud

doi:10.22111/ijfs.2025.50384.8892

Document Type : Research Paper

Authors

¹ semnan university

² Computer and Electrical Eng. Dept. Semnan University, Semnan, Iran

³ Semnan, Iran

⁴ Tehran University

https://doi.org/10.22111/ijfs.2025.50384.8892

Abstract

Message-passing graph neural network (MPNN) is a highly e effective learning tool for graph-structured data but is particularly vulnerable to adversarial attacks on information ow, where poisoned data from attacked nodes can propagate to neighboring nodes and be aggregated, amplifying the negative impact across the graph. The malicious perturbations to node features or graph structures can degrade model performance in defensive decision making, which requires a clean ow. We propose a creative defense mechanism to robustly enhance the resilience of MPGNNs against adversarial perturbations by leveraging a creative knowledge-based fuzzy approach for node classification that is implemented by a rule-based fuzzy model. Our study assigns each neighboring node a membership function, which quantifies its relevance to the target node based on knowledge extracted from structural information, node attributes, and community relationships to the target node. By dynamically adjusting the influence of neighboring nodes, the proposed model mitigates the effects of adversarial noise and reduces the impact of irrelevant or poisoned data, ensuring more robust message aggregation. Furthermore, the fuzzy approach helps prevent over-squashing, a common issue when increasing the number of GNN layers, by maintaining important information through controlled message passing and is followed by picking an opinion leader agent from each egonet to recycle the graph. However, our model does not require prior knowledge of the graph or attack types, making it adaptable to various datasets and attack scenarios. Extensive experiments on benchmark datasets, including Cora, PubMed, and Citeseer, demonstrate greater robustness and better accuracy when compared to state-of-the-art defense strategies with reasonable complexity.

Keywords

Main Subjects

Fuzzy modelling

References

[1] S. Akansha, Over-squashing in graph neural networks: A comprehensive survey, ArXiv preprint arXiv:2308.15568,
(2023). https://doi.org/10.48550/arXiv.2308.15568

[2] I. Alarab, S. Prakoonwit, Adversarial attack for uncertainty estimation: Identifying critical regions in neural
networks, Neural Processing Letters, 54(3) (2011), 1805-1821. https://doi.org/10.1007/s11063-021-10707-3

[3] A. Aleahmad, et al., OLFinder: Finding opinion leaders in online social networks, Journal of Information Science,
42(5) (2016), 659-674. https://doi.org/10.1177/01655515156052

[4] U. Alon, E. Yahav, On the bottleneck of graph neural networks and its practical implications, ArXiv preprint
arXiv:2006.05205, (2020). https://doi.org/10.48550/arXiv.2006.05205

[5] H. Askr, et al., Deep learning in drug discovery: An integrative review and future challenges, Artificial Intelligence
Review, 56(7) (2023), 5975-6037. https://doi.org/10.1007/s10462-022-10306-1

[6] S. Bastami, M. B. Dowlatshahi, Multi-label node classification inheterogeneous networks using graph convolutional
networks, Journal of Mahani Mathematical Research, no. Online First, (2025). https://doi.org/10.22103/jmmr.
2024.23506.1660

[7] E. Bastami, A. Mahabadi, E. Taghizadeh, A gravitation-based link prediction approach in social networks, Swarm
and Evolutionary Computation, 44 (2019), 176-186. doi:https://doi.org/10.1016/j.swevo.2018.03.001

[8] H. Baziyad, V. Kayvanfar, M. Toloo, A data envelopment analysis model for opinion leaders’ identification in
social networks, Computers and Industrial Engineering, 190 (2024), 110010. https://doi.org/10.1016/j.cie.
2024.110010

[9] B. Chamberlain, et al., Grand: Graph neural diffusion, In International Conference on Machine Learning, (2021),
1407-1418. https://doi.org/10.48550/arXiv.2106.10934

[10] D. Chen, Y. Lin, W. Li, P. Li, J. Zhou, X. Sun, Measuring and relieving the over-smoothing problem for graph
neural networks from the topological view, In Proceedings of the AAAI Conference on Artificial Intelligence, 34(0)
(2020), 3438-3445. https://doi.org/10.1609/aaai.v34i04.5747

[11] X. Chen, M. Sim, D. Simchi-levi, P. Sun, Risk aversion in inventory management, Operations Research, 55(5)
(2007), 828-842. https://doi.org/10.1287/opre.1070.0429

[12] N. Drenkow, et al., A systematic review of robustness in deep learning for computer vision: Mind the gap?, ArXiv
preprint arXiv:2112.00639, (2021). https://doi.org/10.48550/arXiv.2112.00639

[13] V. P. Dwivedi, et al., Graph neural networks with learnable structural and positional representations, ArXiv preprint
arXiv:2110.07875, (2021). https://doi.org/10.48550/arXiv.2110.07875

[14] N. Entezari, et al, All you need is low (rank) defending against adversarial attacks on graphs, In Proceedings of
the 13th International Conference on Web Search and Data Mining, (2020), 169-177. https://doi.org/10.1145/
3336191.3371789

[15] S. Geisler, et al., Robustness of graph neural networks at scale, Advances in Neural Information Processing Systems,
34 (2021), 7637-7649. https://doi.org/10.48550/arXiv.2110.14038

[16] S. Geisler, et al., Generalization of neural combinatorial solvers through the lens of adversarial robustness, ArXiv
preprint arXiv:2110.10942, (2021). https://doi.org/10.48550/arXiv.2110.10942

[17] J. Gilmer, et al., Neural message passing for quantum chemistry, International Conference on Machine Learning,
(2017), 1263-1272. https://dl.acm.org/doi/10.5555/3305381.3305512

[18] J. H. Giraldo, et al., On the trade-off between over-smoothing and over-squashing in deep graph neural networks,
In Proceedings of the 32nd ACM International Conference on Information and Knowledge Management, (2023),
566-576. https://doi.org/10.1145/3583780.3614997

[19] A. Goodge, et al., Lunar: Unifying local outlier detection methods via graph neural networks, In Proceedings of the
AAAI Conference on Artificial Intelligence, 36(6) (2022), 6737-6745. https://doi.org/10.1609/aaai.v36i6.
20629

[20] L. Gosch, et al., Adversarial training for graph neural networks: Pitfalls, solutions, and new directions, Advances
in Neural Information Processing Systems, 36 (2024). https://doi.org/10.48550/arXiv.2306.15427

[21] L. Jain, et al., Opinion leaders for information diffusion using graph neural network in online social networks,
ACM Transactions on the Web, 17(2) (2023), 1-37. https://doi.org/10.1145/3580516

[22] W. Jin, et al., Graph structure learning for robust graph neural networks, In Proceedings of the 26th ACM SIGKDD
International Conference on Knowledge Discovery and Data Mining, (2020). https://doi.org/10.1145/3394486.
3403049

[23] W. Jin, et al., Empowering graph representation learning with test-time graph transformation, ArXiv preprint
arXiv:2210.03561, (2022). https://doi.org/10.48550/arXiv.2210.03561

[24] M. Jin, et al., A survey on graph neural networks for time series: Forecasting, classification, imputation, and
anomaly detection, IEEE Transactions on Pattern Analysis and Machine Intelligence, 14(8) (2024), 1-37. https:
//doi.org/10.1109/TPAMI.2024.3443141

[25] T. N. Kipf, M. Welling, Semi-supervised classification with graph convolutional networks, ArXiv preprint
arXiv:1609.02907, (2016). https://doi.org/10.48550/arXiv.1609.02907

[26] S. Kumar, et al., Community detection in complex networks using network embedding and gravitational
search algorithm, Journal of Intelligent Information Systems, 57(1) (2021), 51-72. https://doi.org/10.1007/
s10844-020-00625-6

[27] M. Kumar, et al., Community-enhanced link prediction in dynamic networks, ACM Transactions on the Web, 18(2)
(2024), 1-32. https://doi.org/10.1145/3580513

[28] J. Li, et al., Another perspective of over-smoothing: Alleviating semantic over-smoothing in deep GNNs, IEEE
Transactions on Neural Networks and Learning Systems, (2024), 6897-6910. https://doi.org/10.1109/TNNLS.
2024.3402317

[29] H. Li, et al., Robust knowledge adaptation for dynamic graph neural networks, IEEE Transactions on Knowledge
and Data Engineering, (2024), 6920-6933. https://doi.org/10.1109/TKDE.2024.3388453

[30] F. Li, et al., AdaRisk: Risk-adaptive deep reinforcement learning for vulnerable nodes detection, In IEEE Transactions
on Knowledge and Data Engineering, (2024), 1-14. https://doi.org/10.1109/TKDE.2024.3409869

[31] Y. N. Lin, et al., Fuzzy neural network for representation learning on uncertain graphs, IEEE Transactions on
Fuzzy Systems, (2024), 5259-5271. https://doi.org/10.1109/TFUZZ.2024.3418902

[32] Y. Liu, et al., Explaining GNN over evolving graphs using information flow, ArXiv preprint arXiv:2111.10037,
(2021). https://doi.org/10.48550/arXiv.2111.10037

[33] Y. Liu, M. Liu, X. Xu, Adaptive control design for fixed-time synchronization of fuzzy stochastic cellular neural
networks with discrete and distributed delay, Iranian Journal of Fuzzy Systems, 18(6) (2021), 13-28. https://doi.
org/10.22111/ijfs.2021.6330

[34] A. Mahabadi, M. R. Besmi, Risk-aware service level agreement modeling in smart grid, Multimedia Tools and
Applications, 80(1) (2021), 1433-1456. https://doi.org/10.1007/s11042-020-09787-5

[35] A. Mahabadi, M. Hosseini, SLPA-based parallel overlapping community detection approach in large complex
social networks, Multimedia Tools and Applications, 80(5) (2021), 6567-6598. https://doi.org/10.1007/
s11042-020-09993-1

[36] S. K. Maurya, et al., Not all neighbors are friendly: Learning to choose hop features to improve node classification,
In Proceedings of the 31st ACM International Conference on Information and Knowledge Management, (2022).
https://doi.org/10.1145/3511808.3557543

[37] A. K. McCallum, et al., Automating the construction of internet portals with machine learning, Information Retrieval,
3(2) (2000), 127-163. https://doi.org/10.1023/A:1009953814988

[38] A. Mirhoseini, A graph placement methodology for fast chip design, Nature, 594(7862) (2021), 207-212. https:
//doi.org/10.1038/s41586-021-03544-w

[39] N. J. Momtaz, et al., Identifying opinion leaders for marketing by analyzing online social networks, International
Journal of Virtual Communities and Social Networking (IJVCSN), 3(3) (2011), 19-34. https://doi.org/10.4018/
jvcsn.2011010105

[40] K. Nguyen, et al., Revisiting over-smoothing and over-squashing using ollivier-ricci curvature, In International Conference on Machine Learning, PMLR, (2023), 25956-25979. https://dl.acm.org/doi/10.5555/3618408.3619488

[41] R. Panchendrarajan, A. Saxena, Topic-based influential user detection: A survey, Applied Intelligence, 53(5)
(2023), 5998-6024. https://doi.org/10.1007/s10489-022-03831-7

[42] S. M. Sabharwal, N. Aggrawal, A survey on information diffusion over social network with the application on stock
market and its future prospects, Wireless Personal Communications, 130(4) (2023), 2981-3007. https://doi.org/
10.1007/s11277-023-10412-5

[43] J. Samiuddin, An online self-learning graph-based lateral controller for self-driving cars, IEEE Transactions on
Intelligent Vehicles, (2024). https://doi.org/10.1109/TIV.2024.3478052

[44] P. Sen, et al., Collective classification in network data, AI Magazine, 29(3) (2008), 93-93. https://doi.org/10.
1609/aimag.v29i3.2157

[45] Y. Song, et al., On the robustness of graph neural diffusion to topology perturbations, Advances in Neural Information Processing Systems, 35 (2022), 6384-6396. https://dl.acm.org/doi/abs/10.5555/3600270.3600732

[46] S. Stanovic, B. Ga¨uz´ere, L. Brun, Graph neural networks with maximal independent set-based pooling: Mitigating
over-smoothing and over-squashing, Pattern Recognition Letters, 187 (2025), 14-20. https://doi.org/10.1016/
j.patrec.2024.11.004

[47] J. Topping, et al., Understanding over-squashing and bottlenecks on graphs via curvature, ArXiv preprint
arXiv:2111.14522, (2021). https://doi.org/10.48550/arXiv.2111.14522

[48] J. Vassey, et al., E-cigarette brands and social media influencers on instagram: A social network analysis, Tobacco
Control, 32(e2) (2023), e184-e191. https://doi.org/10.1136/tobaccocontrol-2021-057053

[49] P. Velickovic, et al., Graph attention networks, ArXiv preprint arXiv:1710.10903, (2017). https://doi.org/10.
48550/arXiv.1710.10903

[50] S. Wang, et al., Adversarial defense framework for graph neural network, ArXiv preprint arXiv:1905.03679, (2019).
https://doi.org/10.48550/arXiv.1905.03679

[51] B. Wang, et al., Bandits for black-box attacks to graph neural networks with structure perturbation,
arXiv:2205.03546, (2021). https://doi.org/10.48550/arXiv.2205.03546

[52] S. Wu, et al., Graph neural networks in recommender systems: A survey, ACM Computing Surveys, 55(5) (2022),
1-37. https://doi.org/10.1145/3535101

[53] H. Wu, et al., Adversarial examples on graph data: Deep insights into attack and defense, ArXiv preprint
arXiv:1903.01610, (2019). https://doi.org/10.24963/ijcai.2019/669

[54] K. Xu, et al., Topology attack and defense for graph neural networks: An optimization perspective, ArXiv preprint
arXiv:1906.04214, (2019). https://doi.org/10.48550/arXiv.1906.04214

[55] J. Xu, et al., Robustness of deep learning models on graphs: A survey, AI Open, 2 (2021), 69-78. https://doi.
org/10.1016/j.aiopen.2021.05.002

[56] S. Xu, et al., Cognitive fusion of graph neural network and convolutional neural network for enhanced hyperspectral
target detection, IEEE Transactions on Geoscience and Remote Sensing, 62 (2024), 1-15. https://doi.org/10.
1109/tgrs.2024.3392188

[57] K. Yang, et al., Prediction failure risk-aware decision-making for autonomous vehicles on signalized intersections,
IEEE Transactions on Intelligent Transportation Systems, (2023), 12806-12820. https://doi.org/10.1109/TITS.
2023.3288507

[58] H. Yuan, et al., On explainability of graph neural networks via subgraph explorations, In International Conference
on Machine Learning (ICLR), PMLR, (2021). https://doi.org/10.48550/arXiv.2102.05152

[59] X. Zhang, M. Zitnik, Gnnguard: Defending graph neural networks against adversarial attacks, Advances in Neural
Information Processing Systems, 33 (2020), 9263-9275. https://doi.org/10.48550/arXiv.2006.08149

[60] K. Zhao, et al., Adversarial robustness in graph neural networks: A Hamiltonian approach, Advances in Neural
Information Processing Systems, 36 (2024). https://doi.org/10.1145/3394486.340304

[61] D. Zhu, et al., Robust graph convolutional networks against adversarial attacks, In Proceedings of the 25th ACM
SIGKDD International Conference on Knowledge Discovery and Data Mining, (2019). https://doi.org/10.1145/
3292500.3330851

[62] D. Zugner, et al., Adversarial attacks on neural networks for graph data, In Proceedings of the 24th ACM
SIGKDD International Conference on Knowledge Discovery and Data, (2018). https://doi.org/10.1145/
3219819.3220078

[63] D. Zugner, S. Gunnemann, Adversarial attacks on graph neural networks via meta learning, ArXiv preprint
arXiv:1902.08412, (2019). https://doi.org/10.48550/arXiv.1902.08412

knowledge-based Fuzzy Defensive Classi cation Model for Robust MPGNNs

References

References

Volume 22, Issue 2March and April 2025Pages 97-127

Volume 22, Issue 2
March and April 2025
Pages 97-127