Skip to main content

Table 12 Event extraction performance using various lemmatization tools on GE development set

From: BioLemmatizer: a lemmatization tool for morphological processing of biomedical text

Simple Events

   
 

Recall

Precision

F-score

BioLemmatizer

59.21% (656/1108)

77.82% (656/843)

67.25%

morpha

58.94% (653/1108)

77.92% (653/838)

67.11%

GENIA Tagger

58.84% (652/1108)

77.99% (652/836)

67.08%

WordNet

58.84% (652/1108)

77.90% (652/837)

67.04%

MorphAdorner

58.75% (651/1108)

77.87% (651/836)

66.98%

TreeTagger

58.30% (646/1108)

78.21% (646/826)

66.80%

Binding Events

   
 

Recall

Precision

F-score

TreeTagger

24.66% (92/373)

44.66% (92/206)

31.78%

BioLemmatizer

24.93% (93/373)

43.46% (93/214)

31.69%

morpha

24.93% (93/373)

43.46% (93/214)

31.69%

GENIA Tagger

24.93% (93/373)

43.46% (93/214)

31.69%

MorphAdorner

24.93% (93/373)

43.46% (93/214)

31.69%

WordNet

23.32% (87/373)

43.72% (87/199)

30.42%

Regulation Events

   
 

Recall

Precision

F-score

BioLemmatizer

20.26% (357/1762)

43.17% (357/827)

27.58%

morpha

19.47% (343/1762)

42.93% (343/799)

26.79%

GENIA Tagger

19.52% (344/1762)

42.63% (344/807)

26.78%

MorphAdorner

19.47% (343/1762)

42.56% (343/806)

26.71%

WordNet

19.18% (338/1762)

42.89% (338/788)

26.51%

TreeTagger

19.01% (335/1762)

42.41% (335/790)

26.25%