Abstract:Graph convolutional networks (GCNs) have attracted increasing attention in various fields due to their significant capacity to process graph-structured data. Typically, the GCN model and its variants heavily rely on the transmission of node features across the graph structure, which implicitly assumes that the graph structure and node features are consistent, i.e., they carry related information. However, in many real-world networks, node features may unexpectedly mismatch with the structural information. Existing GCNs fail to generalize to inconsistent scenarios and are even outperformed by models that ignore the graph structure or node features. To address this problem, we investigate how to extract representations from both the graph structure and node features. Consequently, we propose the multi-channel graph convolutional network (MCGCN) for graphs with inconsistent structures and features. Specifically, the MCGCN encodes the graph structure and node features using two specific convolution channels to extract two separate specific representations. Additionally, two joint convolution channels are constructed to extract the common information shared by the graph structure and node features. Finally, an attention mechanism is utilized to adaptively learn the importance weights of these channels under the guidance of the node classification task. In this way, our model can handle both consistent and inconsistent scenarios. Extensive experiments on both synthetic and real-world datasets for node classification and recommendation tasks show that our methods, MCGCN-A and MCGCN-I, achieve the best performance on seven out of eight datasets and the second-best performance on the remaining dataset. For simpler graph structures or tasks where the overhead of multiple convolution channels is not justified, traditional single-channel GCN models might be more efficient.

HV-GCN: Hybrid View Feature Fusion for Graph Convolutional Networks

Multigraph Fusion for Dynamic Graph Convolutional Network

Graph Convolutional Network With Self-Augmented Weights for Semi-Supervised Multi-View Learning

An Efficient Multi-View Heterogeneous Hypergraph Convolutional Network for Heterogeneous Information Network Representation Learning

Joint learning of feature and topology for multi-view graph convolutional network

Learnable Graph Convolutional Network and Feature Fusion for Multi-view Learning

Structure fusion based on graph convolutional networks for semi-supervised classification

Heterogeneous graph convolutional network for multi-view semi-supervised classification

Multi-view Graph Convolutional Networks with Differentiable Node Selection

Revisiting multi-view learning: A perspective of implicitly heterogeneous Graph Convolutional Network

Attention-based stackable graph convolutional network for multi-view learning

Heterogeneous graph convolutional network with local influence

Exploiting Node-Feature Bipartite Graph in Graph Convolutional Networks.

AM-GCN: Adaptive Multi-channel Graph Convolutional Networks

HD-GCN:A Hybrid Diffusion Graph Convolutional Network

Hybrid Micro/Macro Level Convolution for Heterogeneous Graph Learning

Multi-Channel Graph Convolutional Networks for Graphs with Inconsistent Structures and Features

MVMA-GCN: Multi-view Multi-Layer Attention Graph Convolutional Networks

Interpretable and Efficient Heterogeneous Graph Convolutional Network

DAGCN: Dual Attention Graph Convolutional Networks