GraphSAGE

GraphSAGE, developed by researchers including William L. Hamilton, is an inductive graph neural network that generates node embeddings via neighbor sampling, enabling scalable learning on large, evolving graphs.