Informative Graph Structure Learning
About
The quality of graph-structured data is fundamental to the success of modern graph analysis techniques such as Graph Neural Networks (GNNs). However, real-world graph data is often suboptimal, suffering from issues such as noise and incomplete connections. Graph Structure Learning (GSL) has emerged as a promising technique that adaptively optimizes node connections. However, we observe that the effectiveness of GSL often comes at the cost of a dramatic expansion in edge count, resulting in significant storage and computational overhead. In this work, we reveal that this limitation stems from the prevalent use of similarity-based edge construction, which predominantly connects highly similar neighbors based on their embeddings, introducing substantial structure redundancy. To address this, we propose a novel Informative Graph Structure Learning method (InGSL), which jointly considers both similarity and diversity in edge construction by incorporating a mutual-information-guided learning strategy. Notably, InGSL serves as a plug-in module that can be seamlessly integrated into existing GSL frameworks. Through extensive experiments on six representative GSL methods, we demonstrate that InGSL achieves significant performance improvements at a reduced number of edges.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Node Classification | Cora | Accuracy85.83 | 583 | |
| Node Classification | Pubmed | Accuracy83 | 363 | |
| Node Classification | Roman-Empire | Accuracy65.51 | 327 | |
| Node Classification | Ogbn-arxiv | Accuracy71.72 | 235 | |
| Node Classification | Citeseer | Mean Accuracy74.35 | 202 | |
| Node-level classification | BlogCatalog | Accuracy0.9579 | 70 | |
| Node Classification | Cora 70% edge reduction level | Accuracy85.07 | 12 | |
| Node Classification | Citeseer 70% edge reduction level | Accuracy73.16 | 12 | |
| Node Classification | Pubmed 70% edge reduction level | Accuracy82.91 | 12 | |
| Node Classification | Blogcatalog 70% edge reduction level | Accuracy0.9524 | 12 |