Skip to main content

Table 4 Micro-averaged performance for the concept indexing task

From: Combining word embeddings to extract chemical and drug entities in biomedical literature

 

Precision (%)

Recall (%)

F1-score (%)

Classic WE + Contextual WE + Medical WE

92.91

92.44

92.67

Rule + Dictionary-based method (León et al. [62])

91.11

92.08

91.59

Contextual WE

91.11

91.93

91.34

Medical WE

92.16

90.15

91.17

Classic WE

92.13

89.34

90.14

CRF + features [6]

82.89

61.84

70.83