TY - GEN
T1 - Dimensionwise Separable 2-D Graph Convolution for Unsupervised and Semi-Supervised Learning on Graphs
AU - Li, Qimai
AU - Zhang, Xiaotong
AU - Liu, Han
AU - Dai, Quanyu
AU - Wu, Xiao Ming
N1 - Funding Information:
We would like to thank the anonymous reviewers for their insightful comments. This research was supported by the General Research Fund No.15222220 funded by the UGC of Hong Kong.
Publisher Copyright:
© 2021 ACM.
PY - 2021/8/14
Y1 - 2021/8/14
N2 - Graph convolutional neural networks (GCN) have been the model of choice for graph representation learning, which is mainly due to the effective design of graph convolution that computes the representation of a node by aggregating those of its neighbors. However, existing GCN variants commonly use 1-D graph convolution that solely operates on the object link graph without exploring informative relational information among object attributes. This significantly limits their modeling capability and may lead to inferior performance on noisy and sparse real-world networks. In this paper, we explore 2-D graph convolution to jointly model object links and attribute relations for graph representation learning. Specifically, we propose a computationally efficient dimensionwise separable 2-D graph convolution (DSGC) for filtering node features. Theoretically, we show that DSGC can reduce intra-class variance of node features on both the object dimension and the attribute dimension to learn more effective representations. Empirically, we demonstrate that by modeling attribute relations, DSGC achieves significant performance gain over state-of-the-art methods for node classification and clustering on a variety of real-world networks. The source code for reproducing the experimental results is available at https://github.com/liqimai/DSGC.
AB - Graph convolutional neural networks (GCN) have been the model of choice for graph representation learning, which is mainly due to the effective design of graph convolution that computes the representation of a node by aggregating those of its neighbors. However, existing GCN variants commonly use 1-D graph convolution that solely operates on the object link graph without exploring informative relational information among object attributes. This significantly limits their modeling capability and may lead to inferior performance on noisy and sparse real-world networks. In this paper, we explore 2-D graph convolution to jointly model object links and attribute relations for graph representation learning. Specifically, we propose a computationally efficient dimensionwise separable 2-D graph convolution (DSGC) for filtering node features. Theoretically, we show that DSGC can reduce intra-class variance of node features on both the object dimension and the attribute dimension to learn more effective representations. Empirically, we demonstrate that by modeling attribute relations, DSGC achieves significant performance gain over state-of-the-art methods for node classification and clustering on a variety of real-world networks. The source code for reproducing the experimental results is available at https://github.com/liqimai/DSGC.
KW - graph convolution
KW - node classification
KW - node clustering
KW - variance reduction
UR - http://www.scopus.com/inward/record.url?scp=85114911310&partnerID=8YFLogxK
U2 - 10.1145/3447548.3467413
DO - 10.1145/3447548.3467413
M3 - Conference article published in proceeding or book
AN - SCOPUS:85114911310
T3 - Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining
SP - 953
EP - 963
BT - KDD 2021 - Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery and Data Mining
PB - Association for Computing Machinery
T2 - 27th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, KDD 2021
Y2 - 14 August 2021 through 18 August 2021
ER -