Table 7.

Graph embedding algorithms evaluation

Model	Micro-F1	Macro-F1	Weighted-F1
DeepWalk45	0.59	0.29	0.50
Node2Vec46	0.60	0.30	0.51
Walklets47	0.75	0.43	0.70
SocioDim48	0.77	0.51	0.76
RandNE49	0.53	0.24	0.44
NetMF50	0.74	0.49	0.73
GLEE42	0.78	0.54	0.77

The goal is to select algorithms that maximize the preservation of our knowledge graph structure and feature information, GLEE performs the best.