Skip to main content

Table 5 Performance for Linguistic Pattern 1

From: Identifying named entities from PubMed® for enriching semantic categories

Headwords

Total

New

Evaluated

Reviewer 1

Reviewer 2

Reviewer 3

Gene

37678

12461

100

91.0%

91.0%

91.0%

Protein

24000

8630

100

91.0%

91.0%

91.0%

Disease

438

163

163

93.9%

94.5%

93.3%

Cell

50

21

21

95.2%

95.2%

95.2%

Cells

565

380

380

97.1%

97.6%

97.4%

  1. Precisions for each annotator are shown for “gene”, “protein”, “disease”, “cell” and “cells”. “Total” means the total number of obtained terms. “New” and “Evaluated” mean the number of terms not in SemCat and the number of evaluated terms by reviewers, respectively.