{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,10,10]],"date-time":"2025-10-10T01:46:25Z","timestamp":1760060785381,"version":"build-2065373602"},"reference-count":45,"publisher":"MDPI AG","issue":"10","license":[{"start":{"date-parts":[[2025,10,8]],"date-time":"2025-10-08T00:00:00Z","timestamp":1759881600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"the Open Research Fund of Fujian Key Laboratory of Financial Information Processing, Putian University","award":["JXJS202507"],"award-info":[{"award-number":["JXJS202507"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Algorithms"],"abstract":"<jats:p>Graph Neural Networks (GNNs) capture complex information in graph-structured data by integrating node features with iterative updates of graph topology. However, they inherently rely on the homophily assumption\u2014that nodes of the same class tend to form edges. In contrast, real-world networks often exhibit heterophilous structures, where edges are frequently formed between nodes of different classes. Consequently, conventional GNNs, which apply uniform smoothing over all nodes, may inadvertently aggregate both task-relevant and task-irrelevant information, leading to suboptimal performance on heterophilous graphs. In this work, we propose TRed-GNN, a novel end-to-end GNN architecture designed to enhance both the performance and robustness of node classification on heterophilous graphs. The proposed approach decomposes the original graph into a task-relevant subgraph and a task-irrelevant subgraph and employs a dual-channel mechanism to independently aggregate features from each topology. To mitigate the interference of task-irrelevant information, we introduce a reverse process mechanism that, without compromising the main task, extracts potentially useful information from the task-irrelevant subgraph while filtering out noise, thereby improving generalization and resilience to perturbations. Theoretical analysis and extensive experiments on multiple real-world datasets demonstrate that TRed-GNN not only achieves superior classification performance compared to existing methods on most benchmarks, but also exhibits strong adaptability and stability under graph structural perturbations and over-smoothing scenarios.<\/jats:p>","DOI":"10.3390\/a18100632","type":"journal-article","created":{"date-parts":[[2025,10,8]],"date-time":"2025-10-08T12:04:52Z","timestamp":1759925092000},"page":"632","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["TRed-GNN: A Robust Graph Neural Network with Task-Relevant Edge Disentanglement and Reverse Process Mechanism"],"prefix":"10.3390","volume":"18","author":[{"given":"Menghui","family":"Xu","sequence":"first","affiliation":[{"name":"School of Artificial Intelligence, Tiangong University, Tianjin 300387, China"}]},{"given":"Yang","family":"Yan","sequence":"additional","affiliation":[{"name":"School of Information Technology and Engineering, Tianjin University of Technology and Education, Tianjin 300222, China"}]},{"given":"Qiuyan","family":"Wang","sequence":"additional","affiliation":[{"name":"School of Computer Science and Technology, Tiangong University, Tianjin 300387, China"},{"name":"Fujian Key Laboratory of Financial Information Processing, Putian University, Putian 351100, China"}]},{"given":"Hanning","family":"Chen","sequence":"additional","affiliation":[{"name":"School of Artificial Intelligence, Tiangong University, Tianjin 300387, China"},{"name":"College of Artificial Intelligence, Tianjin University of Science and Technology, Tianjin 300457, China"}]},{"given":"Zhao","family":"Zhang","sequence":"additional","affiliation":[{"name":"College of Mechanical Engineering, Tianjin University of Science and Technology, Tianjin 300457, China"}]}],"member":"1968","published-online":{"date-parts":[[2025,10,8]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"110202","DOI":"10.1016\/j.knosys.2022.110202","article-title":"AC2CD: An actor\u2013critic architecture for community detection in dynamic social networks","volume":"261","author":"Costa","year":"2023","journal-title":"Knowl.-Based Syst."},{"key":"ref_2","doi-asserted-by":"crossref","unstructured":"Li, D.X., Zhou, P., Zhao, B.W., Su, X.R., Li, G.D., Zhang, J., Hu, P.W., and Hu, L. (2023). Biocaiv: An integrative webserver for motif-based clustering analysis and interactive visualization of biological networks. BMC Bioinform., 24.","DOI":"10.1186\/s12859-023-05574-9"},{"key":"ref_3","first-page":"4880","article-title":"Graph representation learning beyond node and homophily","volume":"35","author":"Li","year":"2022","journal-title":"IEEE Trans. Knowl. Data Eng."},{"key":"ref_4","doi-asserted-by":"crossref","unstructured":"Zheng, Q., and Zhang, Y. (2023, January 17\u201320). Tagnn: Time adjoint graph neural network for traffic forecasting. Proceedings of the International Conference on Database Systems for Advanced Applications, Tianjin, China.","DOI":"10.1007\/978-3-031-30637-2_24"},{"key":"ref_5","unstructured":"Li, W., Wang, C.h., Cheng, G., and Song, Q. (2021). Optimum-statistical Collaboration Towards General and Efficient Black-box Optimization. arXiv."},{"key":"ref_6","unstructured":"Rusch, T.K., Bronstein, M.M., and Mishra, S. (2023). A survey on oversmoothing in graph neural networks. arXiv."},{"key":"ref_7","first-page":"3438","article-title":"Measuring and relieving the over-smoothing problem for graph neural networks from the topological view","volume":"34","author":"Chen","year":"2020","journal-title":"Proc. AAAI Conf. Artif. Intell."},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"459","DOI":"10.1016\/j.neunet.2023.09.048","article-title":"Exploring the role of edge distribution in graph convolutional networks","volume":"168","author":"He","year":"2023","journal-title":"Neural Netw."},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"102817","DOI":"10.1016\/j.displa.2024.102817","article-title":"Label-aware aggregation on heterophilous graphs for node representation learning","volume":"84","author":"Liu","year":"2024","journal-title":"Displays"},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"275","DOI":"10.1109\/TKDE.2023.3283501","article-title":"Neural moderation of ASMR erotica content in social networks","volume":"36","author":"Chen","year":"2023","journal-title":"IEEE Trans. Knowl. Data Eng."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"11345","DOI":"10.1109\/TPAMI.2024.3459932","article-title":"ES-GNN: Generalizing graph neural networks beyond homophily with edge splitting","volume":"46","author":"Guo","year":"2024","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"ref_12","unstructured":"Kipf, T. (2016). Semi-Supervised Classification with Graph Convolutional Networks. arXiv."},{"key":"ref_13","unstructured":"Veli\u010dkovi\u0107, P., Cucurull, G., Casanova, A., Romero, A., Lio, P., and Bengio, Y. (2017). Graph attention networks. arXiv."},{"key":"ref_14","unstructured":"Hamilton, W., Ying, Z., and Leskovec, J. (2017, January 4\u20139). Inductive representation learning on large graphs. Proceedings of the 31st International Conference on Neural Information Processing Systems, Long Beach, CA, USA."},{"key":"ref_15","unstructured":"Bruna, J., Zaremba, W., Szlam, A., and LeCun, Y. (2013). Spectral networks and locally connected networks on graphs. arXiv."},{"key":"ref_16","unstructured":"Gilmer, J., Schoenholz, S.S., Riley, P.F., Vinyals, O., and Dahl, G.E. (2017, January 6\u201311). Neural message passing for quantum chemistry. Proceedings of the International Conference on Machine Learning, Sydney, NSW, Australia. PMLR."},{"key":"ref_17","first-page":"14239","article-title":"Bernnet: Learning arbitrary graph spectral filters via bernstein approximation","volume":"34","author":"He","year":"2021","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_18","unstructured":"Defferrard, M., Bresson, X., and Vandergheynst, P. (2016, January 5\u201310). Convolutional neural networks on graphs with fast localized spectral filtering. Proceedings of the 30th International Conference on Neural Information Processing System, Barcelona, Spain."},{"key":"ref_19","unstructured":"Pei, H., Wei, B., Chang, K.C.C., Lei, Y., and Yang, B. (2020). Geom-gcn: Geometric graph convolutional networks. arXiv."},{"key":"ref_20","first-page":"7793","article-title":"Beyond homophily in graph neural networks: Current limitations and effective designs","volume":"33","author":"Zhu","year":"2020","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_21","first-page":"1362","article-title":"Revisiting heterophily for graph neural networks","volume":"35","author":"Luan","year":"2022","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_22","doi-asserted-by":"crossref","unstructured":"Wang, R., Mou, S., Wang, X., Xiao, W., Ju, Q., Shi, C., and Xie, X. (2021, January 19\u201323). Graph structure estimation neural networks. Proceedings of the Web Conference 2021, Ljubljana, Slovenia.","DOI":"10.1145\/3442381.3449952"},{"key":"ref_23","first-page":"30414","article-title":"Infogcl: Information-aware graph contrastive learning","volume":"34","author":"Xu","year":"2021","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_24","first-page":"4165","article-title":"Graph structure learning with variational information bottleneck","volume":"36","author":"Sun","year":"2022","journal-title":"Proc. AAAI Conf. Artif. Intell."},{"key":"ref_25","doi-asserted-by":"crossref","unstructured":"Yang, M., Shen, Y., Qi, H., and Yin, B. (2021, January 19\u201323). Soft-mask: Adaptive substructure extractions for graph neural networks. Proceedings of the Web Conference 2021, Ljubljana, Slovenia.","DOI":"10.1145\/3442381.3449929"},{"key":"ref_26","unstructured":"Zheng, C., Zong, B., Cheng, W., Song, D., Ni, J., Yu, W., Chen, H., and Wang, W. (2020, January 13\u201318). Robust graph representation learning via neural sparsification. Proceedings of the International Conference on Machine Learning, Virtual Event. PMLR."},{"key":"ref_27","doi-asserted-by":"crossref","unstructured":"Luo, D., Cheng, W., Yu, W., Zong, B., Ni, J., Chen, H., and Zhang, X. (2021, January 8\u201312). Learning to drop: Robust graph neural network via topological denoising. Proceedings of the 14th ACM International Conference on Web Search and Data Mining, Virtual Event.","DOI":"10.1145\/3437963.3441734"},{"key":"ref_28","unstructured":"Wang, H., and Leskovec, J. (2020). Unifying graph convolutional neural networks and label propagation. arXiv."},{"key":"ref_29","first-page":"76737","article-title":"Interpretable prototype-based graph information bottleneck","volume":"36","author":"Seo","year":"2023","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_30","unstructured":"Higgins, I., Amos, D., Pfau, D., Racaniere, S., Matthey, L., Rezende, D., and Lerchner, A. (2018). Towards a definition of disentangled representations. arXiv."},{"key":"ref_31","first-page":"4916","article-title":"Independence promoted graph disentangled networks","volume":"34","author":"Liu","year":"2020","journal-title":"Proc. AAAI Conf. Artif. Intell."},{"key":"ref_32","unstructured":"Ma, J., Cui, P., Kuang, K., Wang, X., and Zhu, W. (2019, January 9\u201315). Disentangled graph convolutional networks. Proceedings of the International Conference on Machine Learning, Long Beach, CA, USA. PMLR."},{"key":"ref_33","first-page":"20286","article-title":"Factorizable graph convolutional networks","volume":"33","author":"Yang","year":"2020","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_34","doi-asserted-by":"crossref","first-page":"7856","DOI":"10.1109\/TKDE.2022.3206875","article-title":"Disentangled graph contrastive learning with independence promotion","volume":"35","author":"Li","year":"2022","journal-title":"IEEE Trans. Knowl. Data Eng."},{"key":"ref_35","first-page":"11168","article-title":"Graph neural networks with heterophily","volume":"35","author":"Zhu","year":"2021","journal-title":"Proc. AAAI Conf. Artif. Intell."},{"key":"ref_36","doi-asserted-by":"crossref","first-page":"101695","DOI":"10.1016\/j.jocs.2022.101695","article-title":"Simplifying approach to node classification in graph neural networks","volume":"62","author":"Maurya","year":"2022","journal-title":"J. Comput. Sci."},{"key":"ref_37","first-page":"10909","article-title":"Predicting global label relationship matrix for graph neural networks under heterophily","volume":"36","author":"Liang","year":"2023","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_38","unstructured":"Song, Y., Zhou, C., Wang, X., and Lin, Z. (2023). Ordered gnn: Ordering message passing to deal with heterophily and over-smoothing. arXiv."},{"key":"ref_39","first-page":"3950","article-title":"Beyond low-frequency information in graph convolutional networks","volume":"35","author":"Bo","year":"2021","journal-title":"Proc. AAAI Conf. Artif. Intell."},{"key":"ref_40","unstructured":"Chamberlain, B., Rowbottom, J., Gorinova, M.I., Bronstein, M., Webb, S., and Rossi, E. (2021, January 18\u201324). Grand: Graph neural diffusion. Proceedings of the International Conference on Machine Learning, Virtual. PMLR."},{"key":"ref_41","unstructured":"Rong, Y., Huang, W., Xu, T., and Huang, J. (2019). Dropedge: Towards deep graph convolutional networks on node classification. arXiv."},{"key":"ref_42","unstructured":"Wu, F., Souza, A., Zhang, T., Fifty, C., Yu, T., and Weinberger, K. (2019, January 9\u201315). Simplifying graph convolutional networks. Proceedings of the International Conference on Machine Learning, Long Beach, CA, USA. PMLR."},{"key":"ref_43","unstructured":"Gasteiger, J., Bojchevski, A., and G\u00fcnnemann, S. (2018). Predict then propagate: Graph neural networks meet personalized pagerank. arXiv."},{"key":"ref_44","unstructured":"Chien, E., Peng, J., Li, P., and Milenkovic, O. (2020). Adaptive universal generalized pagerank graph neural network. arXiv."},{"key":"ref_45","unstructured":"Abu-El-Haija, S., Perozzi, B., Kapoor, A., Alipourfard, N., Lerman, K., Harutyunyan, H., Ver Steeg, G., and Galstyan, A. (2019, January 9\u201315). Mixhop: Higher-order graph convolutional architectures via sparsified neighborhood mixing. Proceedings of the International Conference on Machine Learning, Long Beach, CA, USA. PMLR."}],"container-title":["Algorithms"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1999-4893\/18\/10\/632\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,9]],"date-time":"2025-10-09T18:50:27Z","timestamp":1760035827000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1999-4893\/18\/10\/632"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,10,8]]},"references-count":45,"journal-issue":{"issue":"10","published-online":{"date-parts":[[2025,10]]}},"alternative-id":["a18100632"],"URL":"https:\/\/doi.org\/10.3390\/a18100632","relation":{},"ISSN":["1999-4893"],"issn-type":[{"type":"electronic","value":"1999-4893"}],"subject":[],"published":{"date-parts":[[2025,10,8]]}}}