Understanding GNNs and Homophily in Dynamic Node Classification

Michael Ito, Danai Koutra, Jenna Wiens
Proceedings of The 28th International Conference on Artificial Intelligence and Statistics, PMLR 258:2764-2772, 2025.

Abstract

Homophily, as a measure, has been critical to increasing our understanding of graph neural networks (GNNs). However, to date this measure has only been analyzed in the context of static graphs. In our work, we explore homophily in dynamic settings. Focusing on graph convolutional networks (GCNs), we demonstrate theoretically that in dynamic settings, current GCN discriminative performance is characterized by the probability that a node’s future label is the same as its neighbors’ current labels. Based on this insight, we propose dynamic homophily, a new measure of homophily that applies in the dynamic setting. This new measure correlates with GNN discriminative performance and sheds light on how to potentially design more powerful GNNs for dynamic graphs. Leveraging a variety of dynamic node classification datasets, we demonstrate that popular GNNs are not robust to low dynamic homophily. Going forward, our work represents an important step towards understanding homophily and GNN performance in dynamic node classification.

Cite this Paper

BibTeX
@InProceedings{pmlr-v258-ito25b, title = {Understanding GNNs and Homophily in Dynamic Node Classification}, author = {Ito, Michael and Koutra, Danai and Wiens, Jenna}, booktitle = {Proceedings of The 28th International Conference on Artificial Intelligence and Statistics}, pages = {2764--2772}, year = {2025}, editor = {Li, Yingzhen and Mandt, Stephan and Agrawal, Shipra and Khan, Emtiyaz}, volume = {258}, series = {Proceedings of Machine Learning Research}, month = {03--05 May}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v258/main/assets/ito25b/ito25b.pdf}, url = {https://proceedings.mlr.press/v258/ito25b.html}, abstract = {Homophily, as a measure, has been critical to increasing our understanding of graph neural networks (GNNs). However, to date this measure has only been analyzed in the context of static graphs. In our work, we explore homophily in dynamic settings. Focusing on graph convolutional networks (GCNs), we demonstrate theoretically that in dynamic settings, current GCN discriminative performance is characterized by the probability that a node’s future label is the same as its neighbors’ current labels. Based on this insight, we propose dynamic homophily, a new measure of homophily that applies in the dynamic setting. This new measure correlates with GNN discriminative performance and sheds light on how to potentially design more powerful GNNs for dynamic graphs. Leveraging a variety of dynamic node classification datasets, we demonstrate that popular GNNs are not robust to low dynamic homophily. Going forward, our work represents an important step towards understanding homophily and GNN performance in dynamic node classification.} }
Endnote
%0 Conference Paper %T Understanding GNNs and Homophily in Dynamic Node Classification %A Michael Ito %A Danai Koutra %A Jenna Wiens %B Proceedings of The 28th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2025 %E Yingzhen Li %E Stephan Mandt %E Shipra Agrawal %E Emtiyaz Khan %F pmlr-v258-ito25b %I PMLR %P 2764--2772 %U https://proceedings.mlr.press/v258/ito25b.html %V 258 %X Homophily, as a measure, has been critical to increasing our understanding of graph neural networks (GNNs). However, to date this measure has only been analyzed in the context of static graphs. In our work, we explore homophily in dynamic settings. Focusing on graph convolutional networks (GCNs), we demonstrate theoretically that in dynamic settings, current GCN discriminative performance is characterized by the probability that a node’s future label is the same as its neighbors’ current labels. Based on this insight, we propose dynamic homophily, a new measure of homophily that applies in the dynamic setting. This new measure correlates with GNN discriminative performance and sheds light on how to potentially design more powerful GNNs for dynamic graphs. Leveraging a variety of dynamic node classification datasets, we demonstrate that popular GNNs are not robust to low dynamic homophily. Going forward, our work represents an important step towards understanding homophily and GNN performance in dynamic node classification.
APA
Ito, M., Koutra, D. & Wiens, J.. (2025). Understanding GNNs and Homophily in Dynamic Node Classification. Proceedings of The 28th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 258:2764-2772 Available from https://proceedings.mlr.press/v258/ito25b.html.

Related Material