BMC Bioinformatics

Table 4 Performance comparison of disease normalization models using various parameters

From: A method for named entity normalization in biomedical articles: application to diseases and plants

				Development set			Test set
Parameters	Win	Dim	Method	Precision	Recall	F-score	Precision	Recall	F-score
5_200_CBOW	5	200	CBOW	0.740	0.918	0.819	0.684	0.896	0.776
5_200_skip	5	200	Skip-gram	0.730	0.909	0.809	0.719	0.890	0.795
5_300_CBOW	5	300	CBOW	0.738	0.918	0.818	0.674	0.892	0.767
5_300_skip	5	300	Skip-gram	0.746	0.916	0.822	0.722	0.893	0.798
5_400_CBOW	5	400	CBOW	0.730	0.918	0.813	0.661	0.878	0.754
5_400_skip	5	400	Skip-gram	0.732	0.916	0.813	0.730	0.905	0.808
7_200_CBOW	7	200	CBOW	0.738	0.919	0.819	0.676	0.891	0.769
7_200_skip	7	200	Skip-gram	0.719	0.900	0.799	0.698	0.882	0.780
7_300_CBOW	7	300	CBOW	0.709	0.911	0.798	0.662	0.880	0.756
7_300_skip	7	300	Skip-gram	0.683	0.895	0.775	0.776	0.769	0.772
7_400_CBOW	7	400	CBOW	0.702	0.898	0.788	0.632	0.850	0.725
7_400_skip	7	400	Skip-gram	0.690	0.896	0.779	0.667	0.887	0.761
8_200_CBOW	8	200	CBOW	0.710	0.907	0.797	0.706	0.891	0.788

The bold font denotes the best result for each column

Back to article page

ISSN: 1471-2105

Contact us

General enquiries: journalsubmissions@springernature.com