{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,13]],"date-time":"2026-04-13T19:01:06Z","timestamp":1776106866728,"version":"3.50.1"},"reference-count":34,"publisher":"Association for Computing Machinery (ACM)","issue":"6","funder":[{"DOI":"10.13039\/100000001","name":"U.S. National Science Foundation","doi-asserted-by":"crossref","award":["IIS-2302786, IIS-2236579, IOS-2430224"],"award-info":[{"award-number":["IIS-2302786, IIS-2236579, IOS-2430224"]}],"id":[{"id":"10.13039\/100000001","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":["ACM Trans. Knowl. Discov. Data"],"published-print":{"date-parts":[[2025,7,31]]},"abstract":"<jats:p>Graph Neural Networks (GNNs) have shown superb performance in handling networked data, mainly attributed to their message passing and convolution process across neighbors. For most literature, the performance of GNNs is mainly reported based on noise-free data environments. No study has systematically evaluated GNNs\u2019 performance under noise. In this article, we carry out an empirical study and theoretical analysis of four types of GNNs, including Graph Convolutional Networks (GCNs), Graph Attention Networks (GATs), Graph Contrastive Networks (GCL), and graph UniFilter under three types of noise, including attribute noise, structure noise, and label noise. Our study shows that GNNs behave tremendously differently in response to different types of noise. Overall, GAT is the most noise vulnerable and sensitive, whereas GCL is the most noise resilient. We further carry out theoretical analysis to explain the reason causing GAT to be sensitive to noise, and propose a solution to enhance its noise resilience. Our study brings in-depth firsthand knowledge of GNNs under noise for researchers and practitioners to better utilize GNNs in real-world applications.<\/jats:p>","DOI":"10.1145\/3733605","type":"journal-article","created":{"date-parts":[[2025,5,1]],"date-time":"2025-05-01T08:37:56Z","timestamp":1746088676000},"page":"1-20","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":3,"title":["A Systematic Study and Analysis of Graph Neural Networks under Noise"],"prefix":"10.1145","volume":"19","author":[{"ORCID":"https:\/\/orcid.org\/0009-0006-6570-123X","authenticated-orcid":false,"given":"Yufei","family":"Jin","sequence":"first","affiliation":[{"name":"Department of Electrical Engineering and Computer Science, Florida Atlantic University, Boca Raton, Florida, USA"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4129-9611","authenticated-orcid":false,"given":"Xingquan","family":"Zhu","sequence":"additional","affiliation":[{"name":"Department of Electrical Engineering and Computer Science, Florida Atlantic University, Boca Raton, Florida, USA"}]}],"member":"320","published-online":{"date-parts":[[2025,6,19]]},"reference":[{"key":"e_1_3_1_2_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01617"},{"key":"e_1_3_1_3_2","doi-asserted-by":"publisher","DOI":"10.1145\/3447548.3467364"},{"key":"e_1_3_1_4_2","doi-asserted-by":"crossref","unstructured":"Enyan Dai Wei Jin Hui Liu and Suhang Wang. 2022. Towards robust graph neural networks for noisy graphs with sparse labels. arXiv:2201.00232. Retrieved from https:\/\/arxiv.org\/abs\/2201.00232","DOI":"10.1145\/3488560.3498408"},{"key":"e_1_3_1_5_2","doi-asserted-by":"publisher","DOI":"10.1145\/3488560.3498408"},{"key":"e_1_3_1_6_2","first-page":"8202","volume-title":"Proceedings of the 40th International Conference on Machine LearningProceedings of Machine Learning Research","volume":"202","author":"Dong Mingze","year":"2023","unstructured":"Mingze Dong and Yuval Kluger. 2023. Towards understanding and reducing graph structural noise for GNNs. In Proceedings of the 40th International Conference on Machine Learning. Andreas Krause, Emma Brunskill, Kyunghyun Cho, Barbara Engelhardt, Sivan Sabato, and Jonathan Scarlett(Eds.), Proceedings of Machine Learning Research, Vol. 202, PMLR, 8202\u20138226. Retrieved from https:\/\/proceedings.mlr.press\/v202\/dong23a.html"},{"key":"e_1_3_1_7_2","doi-asserted-by":"publisher","DOI":"10.1145\/3485447.3512183"},{"key":"e_1_3_1_8_2","first-page":"20310","volume-title":"Proceedings of the 41st International Conference on Machine Learning","volume":"815","author":"Huang Keke","year":"2024","unstructured":"Keke Huang, Yu Guang Wang, Ming Li, and Pietro Li\u00f2. 2024. How universal polynomial bases enhance spectral graph neural networks: Heterophily, over-smoothing, and over-squashing. In Proceedings of the 41st International Conference on Machine Learning, Article 815, 20310\u201320330 pages."},{"key":"e_1_3_1_9_2","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v37i4.25553"},{"key":"e_1_3_1_10_2","volume-title":"Proceedings of the 38th Annual AAAI Conference on Artificial Intelligence","author":"Jin Yufei","year":"2024","unstructured":"Yufei Jin, Richard Gao, Yi He, and Xingquan Zhu. 2024. GLDL: Graph label distribution learning. In Proceedings of the 38th Annual AAAI Conference on Artificial Intelligence."},{"key":"e_1_3_1_11_2","unstructured":"Yufei Jin and Xingquan Zhu. 2024. ATNPA: A unified view of oversmoothing alleviation in graph neural networks. arXiv:2405.01663. Retrieved from https:\/\/arxiv.org\/abs\/2405.01663"},{"key":"e_1_3_1_12_2","volume-title":"Proceedings of the International Conference on Learning Representations (ICLR)","author":"Kipf Thomas N.","year":"2017","unstructured":"Thomas N. Kipf and Max Welling. 2017. Semi-supervised classification with graph convolutional networks. In Proceedings of the International Conference on Learning Representations (ICLR)."},{"key":"e_1_3_1_13_2","doi-asserted-by":"publisher","DOI":"10.1109\/TNNLS.2024.3370918"},{"key":"e_1_3_1_14_2","doi-asserted-by":"publisher","DOI":"10.1109\/TNNLS.2024.3371592"},{"key":"e_1_3_1_15_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.neunet.2024.106113"},{"key":"e_1_3_1_16_2","first-page":"528","volume-title":"Advances in Knowledge Discovery and Data Mining","author":"Li Yayong","year":"2021","unstructured":"Yayong Li, Jie Yin, and Ling Chen. 2021. Unified robust training for graph neural networks against label noise. In Advances in Knowledge Discovery and Data Mining. Kamal Karlapalem, Hong Cheng, Naren Ramakrishnan, R. K. Agrawal, P. Krishna Reddy, Jaideep Srivastava, and Tanmoy Chakraborty (Eds.), Springer International Publishing, Cham, 528\u2013540."},{"key":"e_1_3_1_17_2","unstructured":"Hoang NT Choong Jun Jin and Tsuyoshi Murata. 2019. Learning graph neural networks with noisy labels. arXiv:1905.01591. Retrieved from https:\/\/arxiv.org\/abs\/1905.01591"},{"key":"e_1_3_1_18_2","volume-title":"In Proceedings of the 2020 8th International Conference on Learning Representations (ICLR \u201920)","author":"Pei Hongbin","year":"2020","unstructured":"Hongbin Pei, Bingzhe Wei, Kevin Chen Chuan Chang, Yu Lei, and Bo Yang. 2020. GEOM-GCN: Geometric graph convolutional networks. In Proceedings of the 2020 8th International Conference on Learning Representations (ICLR \u201920). ICLR."},{"key":"e_1_3_1_19_2","volume-title":"Proceedings of the 11th International Conference on Learning Representations","author":"Platonov Oleg","year":"2023","unstructured":"Oleg Platonov, Denis Kuznedelev, Michael Diskin, Babenko Artem, and Prokhorenkova Liudmila. 2023. A critical look at the evaluation of GNNs under heterophily: Are we really making progress? In Proceedings of the 11th International Conference on Learning Representations. Retrieved from https:\/\/openreview.net\/forum?id=tJbbQfw-5wv"},{"key":"e_1_3_1_20_2","doi-asserted-by":"publisher","DOI":"10.1145\/3539597.3570369"},{"key":"e_1_3_1_21_2","volume-title":"Proceedings of the International Conference on Learning Representations","author":"Robinson Joshua David","year":"2021","unstructured":"Joshua David Robinson, Ching-Yao Chuang, Suvrit Sra, and Stefanie Jegelka. 2021. Contrastive learning with hard negative samples. In Proceedings of the International Conference on Learning Representations. Retrieved from https:\/\/openreview.net\/forum?id=CR1XOQ0UTh"},{"key":"e_1_3_1_22_2","first-page":"18888","volume-title":"Proceedings of the 39th International Conference on Machine LearningProceedings of Machine Learning Research","volume":"162","author":"Rusch T. Konstantin","year":"2022","unstructured":"T. Konstantin Rusch, Ben Chamberlain, James Rowbottom, Siddhartha Mishra, and Michael Bronstein. 2022. Graph-coupled oscillator networks. In Proceedings of the 39th International Conference on Machine Learning. Kamalika Chaudhuri, Stefanie Jegelka, Le Song, Csaba Szepesvari, Gang Niu, and Sivan Sabato(Eds.), Proceedings of Machine Learning Research, Vol. 162, PMLR, 18888\u201318909. Retrieved from https:\/\/proceedings.mlr.press\/v162\/rusch22a.html"},{"key":"e_1_3_1_23_2","doi-asserted-by":"publisher","DOI":"10.1145\/1557019.1557108"},{"key":"e_1_3_1_24_2","doi-asserted-by":"publisher","DOI":"10.1145\/3336191.3371851"},{"key":"e_1_3_1_25_2","volume-title":"Proceedings of the International Conference on Learning Representations","author":"Veli\u010dkovi\u0107 Petar","year":"2018","unstructured":"Petar Veli\u010dkovi\u0107, Guillem Cucurull, Arantxa Casanova, Adriana Romero, Pietro Li\u00f2, and Yoshua Bengio. 2018. Graph attention networks. In Proceedings of the International Conference on Learning Representations. Retrieved from https:\/\/openreview.net\/forum?id=rJXMpikCZ"},{"key":"e_1_3_1_26_2","volume-title":"Proceedings of the International Conference on Learning Representations","author":"Veli\u010dkovi\u0107 Petar","year":"2019","unstructured":"Petar Veli\u010dkovi\u0107, William Fedus, William L. Hamilton, Pietro Li\u00f2, Yoshua Bengio, and R. Devon Hjelm. 2019. Deep graph infomax. In Proceedings of the International Conference on Learning Representations. Retrieved from https:\/\/openreview.net\/forum?id=rklz9iAcKQ"},{"key":"e_1_3_1_27_2","unstructured":"Guangtao Wang Rex Ying Jing Huang and Jure Leskovec. 2019. Improving graph attention networks with large margin-based constraints. arXiv:1910.11945. Retrieved from https:\/\/arxiv.org\/abs\/1910.11945"},{"key":"e_1_3_1_28_2","unstructured":"Yancheng Wang and Yingzhen Yang. 2022. Bayesian robust graph contrastive learning. arXiv:2205.14109. Retrieved from https:\/\/arxiv.org\/abs\/2205.14109"},{"key":"e_1_3_1_29_2","unstructured":"Jun Xia Haitao Lin Yongjie Xu Lirong Wu Zhangyang Gao Siyuan Li and Stan Z. Li. 2021. Towards robust graph neural networks against label noise. Retrieved from https:\/\/openreview.net\/forum?id=H38f_9b90BO"},{"key":"e_1_3_1_30_2","first-page":"40","volume-title":"Proceedings of the 33rd International Conference on International Conference on Machine Learning (ICML\u201916)","volume":"48","author":"Yang Zhilin","year":"2016","unstructured":"Zhilin Yang, William W. Cohen, and Ruslan Salakhutdinov. 2016. Revisiting semi-supervised learning with graph embeddings. In Proceedings of the 33rd International Conference on International Conference on Machine Learning (ICML\u201916), Vol. 48. JMLR.org, 40\u201348."},{"key":"e_1_3_1_31_2","doi-asserted-by":"publisher","DOI":"10.1109\/TKDE.2023.3271677"},{"key":"e_1_3_1_32_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICDM50108.2020.00088"},{"key":"e_1_3_1_33_2","doi-asserted-by":"publisher","DOI":"10.1007\/s10462-004-0751-8"},{"key":"e_1_3_1_34_2","volume-title":"Proceedings of the ICML Workshop on Graph Representation Learning and Beyond","author":"Zhu Yanqiao","year":"2020","unstructured":"Yanqiao Zhu, Yichen Xu, Feng Yu, Qiang Liu, Shu Wu, and Liang Wang. 2020. Deep graph contrastive representation learning. In Proceedings of the ICML Workshop on Graph Representation Learning and Beyond. Retrieved from http:\/\/arxiv.org\/abs\/2006.04131"},{"key":"e_1_3_1_35_2","doi-asserted-by":"publisher","DOI":"10.1145\/3219819.3220078"}],"container-title":["ACM Transactions on Knowledge Discovery from Data"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3733605","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,6,19]],"date-time":"2025-06-19T14:55:26Z","timestamp":1750344926000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3733605"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,6,19]]},"references-count":34,"journal-issue":{"issue":"6","published-print":{"date-parts":[[2025,7,31]]}},"alternative-id":["10.1145\/3733605"],"URL":"https:\/\/doi.org\/10.1145\/3733605","relation":{},"ISSN":["1556-4681","1556-472X"],"issn-type":[{"value":"1556-4681","type":"print"},{"value":"1556-472X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,6,19]]},"assertion":[{"value":"2024-05-31","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2025-04-20","order":2,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2025-06-19","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}