SynoExtractor: A Novel Pipeline for Arabic Synonym Extraction Using Word2Vec Word Embeddings

Table 8

The best-obtained results from KSUCCA corpus for p@(1–10), MAP, and MAR after fine tuning CBOW and SG training parameters which were obtained with window size = 2, dimensions = 150, and epochs = 50 (baseline results for KSUCCA corpus).

	P@1	P@2	P@3	P@4	P@5	P@6	P@7	P@8	P@9	P@10	MAP	MAR
CBOW	0.4	0.266	0.22	0.183	0.146	0.122	0.1	0.1	0.088	0.086	0.402	0.111
SG	0.266	0.133	0.155	0.116	0.106	0.088	0.083	0.083	0.074	0.066	0.0272	0.087