Scalability and training time are crucial for any graph neural network model processing a
knowledge graph (KG). While partitioning knowledge graphs helps reduce the training time …