Is Homophily a Necessity for Graph Neural Networks?
By Yao Ma, Xiaorui Liu, Neil Shah, Jiliang Tang. International Conference on Learning Representations, 2021
314 citations recorded by the source corpus.
Graph neural networks (GNNs) have shown great prowess in learning representations suitable for numerous graph-based machine learning tasks. When applied to semi-supervised node classification, GNNs are widely believed to work well due to the homophily assumption ("like attracts like"), and fail to generalize to heterophilous graphs where dissimilar nodes connect. Recent works design new architectures to overcome such heterophily-related limitations, citing poor baseline performance and new architecture improvements on a few heterophilous graph benchmark datasets as evidence for this notion. In our experiments, we empirically find that standard graph convolutional networks (GCNs) can actually achieve better performance than such carefully designed methods on some commonly used heterophilous graphs. This motivates us to reconsider whether homophily is truly necessary for good GNN performance. We find that this claim is not quite true, and in fact, GCNs can achieve strong performance on heterophilous graphs under certain conditions. Our work carefully characterizes these conditions, and provides supporting theoretical understanding and empirical observations. Finally, we examine existing heterophilous graphs benchmarks and reconcile how the GCN (under)performs on them based on this understanding.
Subfields: Artificial Intelligence, Information Systems, Statistical and Nonlinear Physics
View Paper: https://www.semanticscholar.org/paper/8a1e3d41ea3d730e562d8c19b2bdb50a23208842
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
In Information Systems
In Statistical and Nonlinear Physics
Citations in the corpus
- Revisiting Heterophily For Graph Neural Networks (2022, background): “Heterophily has received lots of attention recently and an increasing number of models have been put forward to address this problem [45, 28, 31, 8, 44, 43, 32, 16, 24].”