GI-GCN: Global Interacted Graph Convolutional Networks via Dominant Sets for Graph Classification

Lu Bai, Xinya Qin, Lixin Cui, Ming Li, Hangyuan Du, Ziyu Lyu, Xin Jin
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:5427-5441, 2026.

Abstract

Graph Convolutional Networks (GCNs) are defined based on aggregating the information of adjacent nodes, that are usually treated as equally important and may limit the representational power of existing GCNs. To address this shortcoming, we propose a novel Global Interacted Graph Convolutional Network (GI-GCN), that leverages the solution vectors maintained during the iterative updates of the Dominant Set to adaptively characterize the global importance distribution over all nodes. Specifically, at each convolution layer, this distribution is adopted to adaptively modulate the importance weights of node features before performing the local message passing. We show that this convolution strategy can effectively capture the highly correlated information between nonadjacent nodes through the Dominant Set algorithm, not only emphasizing the critical graph-level information but also enhancing the discriminative power of graph representations. Furthermore, we optimize the memory complexity of the framework, significantly reducing the memory overhead associated with the global interaction modeling. Experiments demonstrate the effectiveness of the proposed GI-GCN model.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-bai26i, title = {{GI}-{GCN}: Global Interacted Graph Convolutional Networks via Dominant Sets for Graph Classification}, author = {Bai, Lu and Qin, Xinya and Cui, Lixin and Li, Ming and Du, Hangyuan and Lyu, Ziyu and Jin, Xin}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {5427--5441}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/bai26i/bai26i.pdf}, url = {https://proceedings.mlr.press/v306/bai26i.html}, abstract = {Graph Convolutional Networks (GCNs) are defined based on aggregating the information of adjacent nodes, that are usually treated as equally important and may limit the representational power of existing GCNs. To address this shortcoming, we propose a novel Global Interacted Graph Convolutional Network (GI-GCN), that leverages the solution vectors maintained during the iterative updates of the Dominant Set to adaptively characterize the global importance distribution over all nodes. Specifically, at each convolution layer, this distribution is adopted to adaptively modulate the importance weights of node features before performing the local message passing. We show that this convolution strategy can effectively capture the highly correlated information between nonadjacent nodes through the Dominant Set algorithm, not only emphasizing the critical graph-level information but also enhancing the discriminative power of graph representations. Furthermore, we optimize the memory complexity of the framework, significantly reducing the memory overhead associated with the global interaction modeling. Experiments demonstrate the effectiveness of the proposed GI-GCN model.} }
Endnote
%0 Conference Paper %T GI-GCN: Global Interacted Graph Convolutional Networks via Dominant Sets for Graph Classification %A Lu Bai %A Xinya Qin %A Lixin Cui %A Ming Li %A Hangyuan Du %A Ziyu Lyu %A Xin Jin %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-bai26i %I PMLR %P 5427--5441 %U https://proceedings.mlr.press/v306/bai26i.html %V 306 %X Graph Convolutional Networks (GCNs) are defined based on aggregating the information of adjacent nodes, that are usually treated as equally important and may limit the representational power of existing GCNs. To address this shortcoming, we propose a novel Global Interacted Graph Convolutional Network (GI-GCN), that leverages the solution vectors maintained during the iterative updates of the Dominant Set to adaptively characterize the global importance distribution over all nodes. Specifically, at each convolution layer, this distribution is adopted to adaptively modulate the importance weights of node features before performing the local message passing. We show that this convolution strategy can effectively capture the highly correlated information between nonadjacent nodes through the Dominant Set algorithm, not only emphasizing the critical graph-level information but also enhancing the discriminative power of graph representations. Furthermore, we optimize the memory complexity of the framework, significantly reducing the memory overhead associated with the global interaction modeling. Experiments demonstrate the effectiveness of the proposed GI-GCN model.
APA
Bai, L., Qin, X., Cui, L., Li, M., Du, H., Lyu, Z. & Jin, X.. (2026). GI-GCN: Global Interacted Graph Convolutional Networks via Dominant Sets for Graph Classification. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:5427-5441 Available from https://proceedings.mlr.press/v306/bai26i.html.

Related Material