Robust Graph Structure Learning with the Alignment of Features and Adjacency Matrix

TL;DR

This paper introduces a robust graph structure learning method that aligns feature and adjacency information, improving GNN performance on noisy graphs by leveraging theoretical insights and low-dimensional features.

Contribution

It proposes a novel regularized GSL approach that aligns feature and graph information, supported by a new theoretical lower bound on node-level Rademacher complexity.

Findings

Outperforms baseline methods on real-world noisy graphs

Effective in scenarios with heavily corrupted graph structures

Demonstrates the importance of feature-graph alignment in GSL

Abstract

To improve the robustness of graph neural networks (GNN), graph structure learning (GSL) has attracted great interest due to the pervasiveness of noise in graph data. Many approaches have been proposed for GSL to jointly learn a clean graph structure and corresponding representations. To extend the previous work, this paper proposes a novel regularized GSL approach, particularly with an alignment of feature information and graph information, which is motivated mainly by our derived lower bound of node-level Rademacher complexity for GNNs. Additionally, our proposed approach incorporates sparse dimensional reduction to leverage low-dimensional node features that are relevant to the graph structure. To evaluate the effectiveness of our approach, we conduct experiments on real-world graphs. The results demonstrate that our proposed GSL method outperforms several competitive baselines, especially in scenarios where the graph structures are heavily affected by noise. Overall, our research highlights the importance of integrating feature and graph information alignment in GSL, as inspired by our derived theoretical result, and showcases the superiority of our approach in handling noisy graph structures through comprehensive experiments on real-world datasets.