Is Heterophily A Real Nightmare For Graph Neural Networks To Do Node Classification?
This addresses performance issues in GNNs for node classification under heterophily, offering a novel framework that improves accuracy across multiple real-world datasets, though it is incremental in building on existing GNN methods.
The paper tackles the problem of heterophily in Graph Neural Networks (GNNs) for node classification, showing that not all heterophily cases are harmful and proposing an Adaptive Channel Mixing (ACM) framework that adaptively uses aggregation, diversification, and identity channels, which achieves significant performance gains and exceeds state-of-the-art GNNs on most of 10 real-world tasks without high computational cost.
Graph Neural Networks (GNNs) extend basic Neural Networks (NNs) by using the graph structures based on the relational inductive bias (homophily assumption). Though GNNs are believed to outperform NNs in real-world tasks, performance advantages of GNNs over graph-agnostic NNs seem not generally satisfactory. Heterophily has been considered as a main cause and numerous works have been put forward to address it. In this paper, we first show that not all cases of heterophily are harmful for GNNs with aggregation operation. Then, we propose new metrics based on a similarity matrix which considers the influence of both graph structure and input features on GNNs. The metrics demonstrate advantages over the commonly used homophily metrics by tests on synthetic graphs. From the metrics and the observations, we find some cases of harmful heterophily can be addressed by diversification operation. With this fact and knowledge of filterbanks, we propose the Adaptive Channel Mixing (ACM) framework to adaptively exploit aggregation, diversification and identity channels in each GNN layer to address harmful heterophily. We validate the ACM-augmented baselines with 10 real-world node classification tasks. They consistently achieve significant performance gain and exceed the state-of-the-art GNNs on most of the tasks without incurring significant computational burden.