Skip to main content
. 2021 Aug 30;8:636077. doi: 10.3389/fmolb.2021.636077

TABLE 5.

The 20 most frequent tokens, including stop words and punctuation marks, within various window sizes around entities incorrectly labeled by human reviewers. Words that are neither stop words nor punctuation words are in boldface.

Window size = 1 Window size = 3 Window size = 5
# Token Count Token Count Token Count
1 27 49 74
2 14 And 21 28
3 And 14 Mg 19 And 28
4 with 8 18 Mg 23
5 ( 7 ) 13 ) 21
6 Is 7 With 10 Of 18
7 Of 6 Of 10 ( 17
8 Was 6 ( 9 with 13
9 Or 4 Is 9 to 12
10 300 4 Or 7 the 11
11 oral 3 once/day 7 a 11
12 Has 3 A 7 in 10
13 To 3 The 6 is 9
14 [ 3 Was 6 or 9
15 dose 3 To 6 daily 8
16 intravenous 2 treatment 6 was 8
17 In 2 In 5 treatment 8
18 may 2 300 5 ] 8
19 500 2 Be 4 once/day 8
20 A 2 Treated 4 were 7