Star Topology Convolution for Graph Representation Learning
Preprints are manuscripts made publicly available before they have been submitted for formal peer review and publication. They might contain new research findings or data. Preprints can be a draft or final version of an author's research but must not have been accepted for publication at the time of submission.
We present a novel graph convolutional method called star topology convolution (STC). This method makes graph convolution more similar to conventional convolutional in neural networks (CNNs) in Euclidean feature space. Unlike most existing spectral convolution methods, this method learns subgraphs which have a star topology rather than a fixed graph. It has fewer parameters in its convolution kernel and is inductive so that it is more flexible and can be applied to large and evolving graphs. As for CNNs in Euclidean feature spaces, the convolution kernel is localized and maintains good sharing. By increasing the depth of a layer, the method can learn lobal features like a CNN. To validate the method, STC was compared to state-of-the-art spectral convolution and spatial convolution methods in a supervised learning setting on three benchmark datasets: Cora, Citeseer and Pubmed. The experimental results show that STC outperforms the other methods. STC was also applied to protein identification tasks and outperformed traditional and advanced protein identification methods.