Table 1. Hyperparameter optimization values for determining the best performing knowledge graph embedding model.

Hyperparameter	Values
Batch count	8, 12, 16
Embedding size	90, 120, 150
Number of negatives (eta) per positive triple	10, 15, 20
Loss function	pairwise, multiclass_nll
Regularization type	L1, L2, nuclear 3-norm
Learning rate	0.0001, 0.001