Skip to main content

Table 3 Levels of ambiguity in LINNAEUS species tags on different document sets.

From: LINNAEUS: A species name identification system for biomedical literature

 

None

Earlier

Whole

 

Strict

Approx.

Strict

Approx.

Strict

Approx.

MEDLINE

0.111

0.053

0.059

0.030

0.054

0.028

PMC OA abs

0.110

0.061

0.054

0.031

0.049

0.028

PMC OA

0.143

0.075

0.029

0.015

0.027

0.013

  1. "None" refers to the baseline case where no disambiguation is performed, "earlier" refers to disambiguation of an ambiguous mention by searching for its explicit species mentions earlier in the document and "whole" refers to disambiguation by searching for its explicit mentions across the whole document. In the "approximate" mode, a heuristic is employed to further disambiguate ambiguous mentions based on the probability of correct species usage.