Skip to main content

Table 5 NER performance comparison results of BERT and bi-LSTM-CRF on the diagnosis dataset

From: Korean clinical entity recognition from diagnosis text using BERT

 

BERT (IOB)

bi-LSTM-CRF (IOB)

 

macro-avg

micro-avg

macro-avg

micro-avg

Precision

0.82

0.83

0.81

0.82

Recall

0.85

0.85

0.78

0.79

F1

0.83

0.84

0.79

0.81