A Survey on Graph Structure Learning: Progress and Opportunities

TL;DR

This survey reviews recent advances in Graph Structure Learning (GSL), highlighting methods for jointly optimizing graph structures and representations to improve graph neural network performance across various applications.

Contribution

It provides a comprehensive overview of GSL methods, categorizes them by modeling approaches, and discusses future research directions in the field.

Findings

GSL methods enhance GNN performance on multiple tasks.

Current GSL approaches face challenges like scalability and robustness.

Future directions include addressing these issues and expanding applications.

Abstract

Graphs are widely used to describe real-world objects and their interactions. Graph Neural Networks (GNNs) as a de facto model for analyzing graphstructured data, are highly sensitive to the quality of the given graph structures. Therefore, noisy or incomplete graphs often lead to unsatisfactory representations and prevent us from fully understanding the mechanism underlying the system. In pursuit of an optimal graph structure for downstream tasks, recent studies have sparked an effort around the central theme of Graph Structure Learning (GSL), which aims to jointly learn an optimized graph structure and corresponding graph representations. In the presented survey, we broadly review recent progress in GSL methods. Specifically, we first formulate a general pipeline of GSL and review state-of-the-art methods classified by the way of modeling graph structures, followed by applications of GSL across domains. Finally, we point out some issues in current studies and discuss future directions.