Cross-GCN: Enhancing Graph Convolutional Network with $k$-Order Feature Interactions
This addresses a bottleneck in GCNs for tasks like recommendation and classification where cross features are crucial, offering a domain-specific improvement.
The authors tackled the problem that Graph Convolutional Networks (GCNs) lack explicit modeling of cross features, which reduces effectiveness in tasks where such features are important, by proposing Cross-GCN, a new architecture that explicitly models arbitrary-order cross features with linear complexity, achieving state-of-the-art performance on three graph datasets.
Graph Convolutional Network (GCN) is an emerging technique that performs learning and reasoning on graph data. It operates feature learning on the graph structure, through aggregating the features of the neighbor nodes to obtain the embedding of each target node. Owing to the strong representation power, recent research shows that GCN achieves state-of-the-art performance on several tasks such as recommendation and linked document classification. Despite its effectiveness, we argue that existing designs of GCN forgo modeling cross features, making GCN less effective for tasks or data where cross features are important. Although neural network can approximate any continuous function, including the multiplication operator for modeling feature crosses, it can be rather inefficient to do so (i.e., wasting many parameters at the risk of overfitting) if there is no explicit design. To this end, we design a new operator named Cross-feature Graph Convolution, which explicitly models the arbitrary-order cross features with complexity linear to feature dimension and order size. We term our proposed architecture as Cross-GCN, and conduct experiments on three graphs to validate its effectiveness. Extensive analysis validates the utility of explicitly modeling cross features in GCN, especially for feature learning at lower layers.