🤖 AI Summary
This paper addresses the pervasive “over-compression” problem in Graph Neural Networks (GNNs), wherein multi-hop neighborhood representations become indistinguishable, undermining long-range dependency modeling. We first establish a unified analytical framework that rigorously distinguishes over-compression from over-smoothing. Methodologically, we propose a taxonomy of mitigation strategies—graph rewiring, curvature-aware propagation, and spectral normalization—and integrate discrete curvature modeling, adaptive spectral analysis, and a novel normalization technique. Extensive cross-dataset empirical evaluation is conducted on multiple benchmarks. Our contributions include: (i) a theoretically grounded characterization of over-compression; (ii) a principled, modular framework unifying diverse mitigation approaches; and (iii) a comprehensive knowledge graph spanning theory, model design, and evaluation protocols—including open-source implementations and standardized benchmarks—to advance reproducible research on long-range modeling in GNNs.
📝 Abstract
Graph Neural Networks (GNNs) revolutionize machine learning for graph-structured data, effectively capturing complex relationships. They disseminate information through interconnected nodes, but long-range interactions face challenges known as"over-squashing". This survey delves into the challenge of over-squashing in Graph Neural Networks (GNNs), where long-range information dissemination is hindered, impacting tasks reliant on intricate long-distance interactions. It comprehensively explores the causes, consequences, and mitigation strategies for over-squashing. Various methodologies are reviewed, including graph rewiring, novel normalization, spectral analysis, and curvature-based strategies, with a focus on their trade-offs and effectiveness. The survey also discusses the interplay between over-squashing and other GNN limitations, such as over-smoothing, and provides a taxonomy of models designed to address these issues in node and graph-level tasks. Benchmark datasets for performance evaluation are also detailed, making this survey a valuable resource for researchers and practitioners in the GNN field.