Skip to main content

Table 8 NER performance before and after typo correction in the NCBI-disease dataset

From: MLM-based typographical error correction of unstructured medical texts for named entity recognition

NER

Precision

Recall

f1-score

Support

Typos

Typo correction

Typos

Typo correction

Typos

Typo correction

B-Disease

0.96

0.95

0.73

0.74

0.83

0.84

960

I-Disease

0.97

0.97

0.66

0.71

0.79

0.82

1087

 

Typos

Typo correction

 

Accuracy

0.70

0.72

2047

f1-score

0.77

0.79

  1. Dataset included 16% errors. The two tagged values, B-Disease and I-Disease, appeared 960 and 1087 times, respectively.
Back to article page

BMC Bioinformatics

ISSN: 1471-2105

Contact us