Inductive Representation Learning on Large Graphs

NeurIPS 2017 William L. HamiltonRex YingJure Leskovec

Low-dimensional embeddings of nodes in large graphs have proved extremely useful in a variety of prediction tasks, from content recommendation to identifying protein functions. However, most existing approaches require that all nodes in the graph are present during training of the embeddings; these previous approaches are inherently transductive and do not naturally generalize to unseen nodes... (read more)

PDF Abstract

Evaluation results from the paper


Task Dataset Model Metric name Metric value Global rank Compare
Node Classification CiteSeer (0.5%) GraphSAGE Accuracy 33.8% # 14
Node Classification CiteSeer (1%) GraphSAGE Accuracy 51.0% # 12
Node Classification CiteSeer with Public Split: fixed 20 nodes per class GraphSAGE Accuracy 67.2% # 11
Node Classification Cora (0.5%) GraphSAGE Accuracy 37.5% # 13
Node Classification Cora (1%) GraphSAGE Accuracy 49.0% # 12
Node Classification Cora (3%) GraphSAGE Accuracy 64.2% # 12
Node Classification Cora with Public Split: fixed 20 nodes per class GraphSAGE Accuracy 74.5% # 15
Node Classification PPI GraphSAGE F1 61.2 # 10
Node Classification PubMed (0.03%) GraphSAGE Accuracy 45.4% # 13
Node Classification PubMed (0.05%) GraphSAGE Accuracy 53.0% # 12
Node Classification PubMed (0.1%) GraphSAGE Accuracy 65.4% # 12
Node Classification PubMed with Public Split: fixed 20 nodes per class GraphSAGE Accuracy 76.8% # 10