BiGCN: A Bi-directional Low-Pass Filtering Graph Neural Network
This work addresses robustness issues in graph neural networks for applications like social networks and recommendation systems, but it is incremental as it builds on existing low-pass filtering concepts.
The authors tackled the problem of noise robustness in graph convolutional networks by proposing BiGCN, a bidirectional low-pass filtering graph neural network that incorporates both original graph structure and latent feature correlations, resulting in state-of-the-art performance on node classification and link prediction tasks in citation and co-purchase networks under various noise settings.
Graph convolutional networks (GCNs) have achieved great success on graph-structured data. Many graph convolutional networks can be thought of as low-pass filters for graph signals. In this paper, we propose a more powerful graph convolutional network, named BiGCN, that extends to bidirectional filtering. Specifically, we not only consider the original graph structure information but also the latent correlation between features, thus BiGCN can filter the signals along with both the original graph and a latent feature-connection graph. Compared with most existing GCNs, BiGCN is more robust and has powerful capacities for feature denoising. We perform node classification and link prediction in citation networks and co-purchase networks with three settings: noise-rate, noise-level and structure-mistakes. Extensive experimental results demonstrate that our model outperforms the state-of-the-art graph neural networks in both clean and artificially noisy data.