Eigen-GNN: A Graph Structure Preserving Plug-in for GNNs
This addresses a fundamental challenge in graph machine learning for practitioners using shallow GNNs, offering a general solution to improve structure preservation without increasing model depth.
The paper tackles the problem that shallow Graph Neural Networks (GNNs) fail to preserve graph structures well, proposing Eigen-GNN as a plug-in module that integrates eigenspace to boost structure preservation, with experimental results showing effectiveness in tasks like node classification and link prediction.
Graph Neural Networks (GNNs) are emerging machine learning models on graphs. Although sufficiently deep GNNs are shown theoretically capable of fully preserving graph structures, most existing GNN models in practice are shallow and essentially feature-centric. We show empirically and analytically that the existing shallow GNNs cannot preserve graph structures well. To overcome this fundamental challenge, we propose Eigen-GNN, a simple yet effective and general plug-in module to boost GNNs ability in preserving graph structures. Specifically, we integrate the eigenspace of graph structures with GNNs by treating GNNs as a type of dimensionality reduction and expanding the initial dimensionality reduction bases. Without needing to increase depths, Eigen-GNN possesses more flexibilities in handling both feature-driven and structure-driven tasks since the initial bases contain both node features and graph structures. We present extensive experimental results to demonstrate the effectiveness of Eigen-GNN for tasks including node classification, link prediction, and graph isomorphism tests.