Graph Mediator Networks Bridging Local and Global Semantics via Serial Message Passing

Jiangfeng Sun, SiHao He, Yanlong Lin, Zhonghong Ou, Meina Song
Proceedings of the 17th Asian Conference on Machine Learning, PMLR 304:495-510, 2025.

Abstract

Graph Neural Networks (GNNs) have achieved remarkable success in modeling structured data through local message passing. However, their effectiveness diminishes on graphs with low homophily or irregular structures, where long-range dependencies are hard to capture and features tend to suffer from over-smoothing and noise amplification. To address these limitations, we propose GMN, a novel dual-path Graph Mediator Network that explicitly enhances both global information propagation and spectral stability. In the spatial path, GMN introduces a lightweight Mediator node connected to all graph nodes, allowing long-range communication to occur in a single hop without increasing network depth. In parallel, the spectral path leverages multi-scale Chebyshev filtering along with a spectral energy regularization term that suppresses high-frequency noise, leading to smoother and more stable node embeddings. These two complementary pathways are adaptively integrated via a gated fusion mechanism, which dynamically balances their contributions based on structural context. Final graph-level representations are obtained through task-specific pooling strategies, enabling GMN to generalize effectively across different tasks. Extensive experiments on benchmark datasets with varying homophily levels and structural perturbations demonstrate that GMN consistently achieves state-of-the-art performance in terms of accuracy, robustness, and generalization. Code is available at: https://github.com/sun2017bupt/GMN.

Cite this Paper

BibTeX
@InProceedings{pmlr-v304-sun25a, title = {Graph Mediator Networks Bridging Local and Global Semantics via Serial Message Passing}, author = {Sun, Jiangfeng and He, SiHao and Lin, Yanlong and Ou, Zhonghong and Song, Meina}, booktitle = {Proceedings of the 17th Asian Conference on Machine Learning}, pages = {495--510}, year = {2025}, editor = {Lee, Hung-yi and Liu, Tongliang}, volume = {304}, series = {Proceedings of Machine Learning Research}, month = {09--12 Dec}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v304/main/assets/sun25a/sun25a.pdf}, url = {https://proceedings.mlr.press/v304/sun25a.html}, abstract = {Graph Neural Networks (GNNs) have achieved remarkable success in modeling structured data through local message passing. However, their effectiveness diminishes on graphs with low homophily or irregular structures, where long-range dependencies are hard to capture and features tend to suffer from over-smoothing and noise amplification. To address these limitations, we propose GMN, a novel dual-path Graph Mediator Network that explicitly enhances both global information propagation and spectral stability. In the spatial path, GMN introduces a lightweight Mediator node connected to all graph nodes, allowing long-range communication to occur in a single hop without increasing network depth. In parallel, the spectral path leverages multi-scale Chebyshev filtering along with a spectral energy regularization term that suppresses high-frequency noise, leading to smoother and more stable node embeddings. These two complementary pathways are adaptively integrated via a gated fusion mechanism, which dynamically balances their contributions based on structural context. Final graph-level representations are obtained through task-specific pooling strategies, enabling GMN to generalize effectively across different tasks. Extensive experiments on benchmark datasets with varying homophily levels and structural perturbations demonstrate that GMN consistently achieves state-of-the-art performance in terms of accuracy, robustness, and generalization. Code is available at: https://github.com/sun2017bupt/GMN.} }
Endnote
%0 Conference Paper %T Graph Mediator Networks Bridging Local and Global Semantics via Serial Message Passing %A Jiangfeng Sun %A SiHao He %A Yanlong Lin %A Zhonghong Ou %A Meina Song %B Proceedings of the 17th Asian Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Hung-yi Lee %E Tongliang Liu %F pmlr-v304-sun25a %I PMLR %P 495--510 %U https://proceedings.mlr.press/v304/sun25a.html %V 304 %X Graph Neural Networks (GNNs) have achieved remarkable success in modeling structured data through local message passing. However, their effectiveness diminishes on graphs with low homophily or irregular structures, where long-range dependencies are hard to capture and features tend to suffer from over-smoothing and noise amplification. To address these limitations, we propose GMN, a novel dual-path Graph Mediator Network that explicitly enhances both global information propagation and spectral stability. In the spatial path, GMN introduces a lightweight Mediator node connected to all graph nodes, allowing long-range communication to occur in a single hop without increasing network depth. In parallel, the spectral path leverages multi-scale Chebyshev filtering along with a spectral energy regularization term that suppresses high-frequency noise, leading to smoother and more stable node embeddings. These two complementary pathways are adaptively integrated via a gated fusion mechanism, which dynamically balances their contributions based on structural context. Final graph-level representations are obtained through task-specific pooling strategies, enabling GMN to generalize effectively across different tasks. Extensive experiments on benchmark datasets with varying homophily levels and structural perturbations demonstrate that GMN consistently achieves state-of-the-art performance in terms of accuracy, robustness, and generalization. Code is available at: https://github.com/sun2017bupt/GMN.
APA
Sun, J., He, S., Lin, Y., Ou, Z. & Song, M.. (2025). Graph Mediator Networks Bridging Local and Global Semantics via Serial Message Passing. Proceedings of the 17th Asian Conference on Machine Learning, in Proceedings of Machine Learning Research 304:495-510 Available from https://proceedings.mlr.press/v304/sun25a.html.

Related Material