<ul><li>Graph-structured data is prevalent in various domains like social networks, biological systems, and recommender systems.</li><li>Graph Foundation Models (GFMs) aim to extend the success of foundation models in natural language processing and vision to graphs, which present unique challenges and opportunities due to their non-Euclidean structures.</li><li>The survey provides a comprehensive overview of GFMs, categorizing them based on generalization scope and highlighting key components such as backbone architectures, pretraining strategies, and adaptation mechanisms.</li><li>GFMs are seen as foundational infrastructure for reasoning over structured data, addressing challenges such as structural alignment, heterogeneity, scalability, and evaluation.</li></ul>

Graph Foundation Models: A Comprehensive Survey

Discover more