A corpus of full-text journal articles is a robust evaluation tool for revealing differences in performance of biomedical natural language processing tools

Table 2 Semantic class groupings for CRAFT

Mapping strategy	Source Semantic class
STAR	(any class)
GENE	gene
POLY	polypeptide
	QTL
	cDNA
	gene
GENESTAR	gene or polypeptide
	gene or polypeptide or macromolecular complex
	gene or transcript or polypeptide
	gene or transcript or macromolecular complex
	macromolecular complex
	polypeptide
	polypeptide or macromolecular complex
POLYSTAR	promoter
	transcript
	transcript or polypeptide
	transcript or polypeptide or macromolecular complex

ISSN: 1471-2105