Skip to main content

Table 5 Verb tokens covered by the 29 most frequent verbs in each corpus. These counts reflect filtering some non-biomedical verbs, such as be. Compare these data to those in Table 3.

From: A critical review of PASBio's argument structures for biomedical verbs

Corpus

Percentage

Tokens

BioIE (both)

23.8%

4,088/17,186

BioIE-P450

29.2%

2,757/9,455

BioIE-Oncology

21.7%

1,675/7,731

GENIA

29.6%

15,363/51,879