Table 2.

Results of the application of the extracted patterns

Corpus	#	Recall	Precision	f − s c o r e	f − s c o r e
	Sentences				presented in [ 11 ]
					details given in Table 3
AIMed	1955	78.6	35.6	49	[34.7,41.5]
BioInfer	1100	46.5	25.3	32.8	[15.9,40.6]
PubMed	200	75.0	83.0	78.7	−
HPRD50	145	66.8	46.7	55.0	[38.3,69.8]

Table 2 gives the list of the four testing corpora used to evaluate the proposed approach, and the results of the evaluation. The meaning of the columns is: the name of the corpus, the number of the sentences in the corpus, the recall score of the proposed approach applied on the corpus, the precision score of the proposed approach applied on the corpus, thef-score of the proposed approach applied on the corpus. The last column indicates the range of the f-scores presented in [11] with also a cross-corpus validation.