Abstractive Arabic Text Summarization Based on Deep Learning

Table 6

Performance comparisons of the proposed system using several word embedding models.

Dataset	word2vec	ROUGE-1			ROUGE-2			ROUGE-L			BLEU
		Precision	Recall	F1	Precision	Recall	F1	Precision	Recall	F1
AHS	Skip-gram	54.95	50.48	51.49	13.1	12.01	12.27	37.84	35.19	34.37	0.41
	CBOW	52.25	48.22	49.08	11.11	10.26	10.46	36.13	33.73	32.85	0.37
AMN	Skip-gram	48.15	42.65	44.28	19.46	17.93	18.35	35.48	32.86	32.46	0.41
	CBOW	43.78	38.83	40.11	17.05	15.21	15.92	32.28	29.7	29.03	0.38