Modeling Relational Data with Graph Convolutional Networks

Benchmark Model Rank Results
heterogeneous-node-classification-on-acmRGCN#6Macro-F1: 91.55Micro-F1: 91.41
heterogeneous-node-classification-on-dblp-2RGCN#7Macro-F1: 91.52Micro-F1: 92.07
heterogeneous-node-classification-on-freebaseRGCN#5Macro-F1: 46.78Micro-F1: 58.33
heterogeneous-node-classification-on-imdbRGCN#7Macro-F1: 58.85Micro-F1: 62.05
heterogeneous-node-classification-on-oagRGCN#3NDCG: 48.93MRR: 31.51
heterogeneous-node-classification-on-oag-l1RGCN#4NDCG: 85.91MRR: 84.92
node-classification-on-aifbR-GCN#1Accuracy: 95.83
node-classification-on-amR-GCN#4Accuracy: 89.29
node-classification-on-bgsR-GCN#6Accuracy: 83.10
node-classification-on-mutagR-GCN#5Accuracy: 73.23
node-property-prediction-on-ogbn-magR-GSN#22Test Accuracy: 0.5032 ± 0.0037Ext. data: No
node-property-prediction-on-ogbn-magFull-batch R-GCN#30Test Accuracy: 0.3977 ± 0.0046Ext. data: No