Table 4.

eFIP performance evaluation on information extraction at the sentence level as measured by precision, recall and F-measure based on TP, TN, FP and FN

Evaluation type	# Abstracts/sentences	Time to completion	Precision	Recall	F-measure	TP	TN	FP	FN
System evaluation (In-house)	96/148		72.4	67.9	70.1	55	46	21	26
User evaluation (BioCreative)
Set 1: Manual curation	25/37	104 min	84.2	80.0	82.0	16	14	3	4
Set 2: eFIP interface	25/37	42.5 min	94.7	69.2	80.0	18	10	1	8