Inductive Representation Learning on Large Graphs (GraphSAGE)¶
Authors: William L. Hamilton, Rex Ying, Jure Leskovec Year: 2017 Citations: 15,000+
Summary¶
GraphSAGE (Sample and Aggregate) enables inductive GNN learning on unseen nodes. Samples and aggregates from neighborhoods instead of computing all node embeddings.
Key Concepts¶
- Inductive Learning: Learn embeddings for unseen nodes
- Neighborhood Sampling: Sample and aggregate K-hop neighbors
- Aggregation Functions: Mean, LSTM, pooling combine features
- Mini-Batch Training: Efficient sampling-based training
- Scalability: Works on massive graphs beyond memory
- Generalization: Learned aggregations transfer to new nodes
Impact¶
- 15,000+ citations
- Enabled inductive learning on graphs
- More practical than transductive GCN
- Better scalability for production systems
- Influenced subsequent GNN designs
- Foundation for production GNNs
- Applications in recommendation, social networks
Related Papers¶
- Graph Convolutional Networks (Kipf & Welling, 2017)
- Graph Attention Networks (Veličković et al., 2017)
- Node2Vec (Grover & Leskovec, 2016)