Table 5 Sentence pair similarity results on BioCreative/OHNLP STS dataset.

From: BioWordVec, improving biomedical word embeddings with subword information and MeSH

Similarity measures

Mikolov et al.1

Pyysalo et al.21

Chiu et al.8

BioWordVec (win20) w/o MeSH

BioWordVec (win20) w/MeSH

Cosine

0.768

0.755

0.757

0.770

0.771

Euclidean

0.725

0.723

0.727

0.751

0.753

Block

0.725

0.722

0.727

0.750

0.752

  1. “win20” denotes the BioWordVec was trained by setting the context window size as 20.