Skip to main content

Table 1 Datasets used in our experiment

From: Exploiting likely-positive and unlabeled data to improve the identification of protein-protein interaction articles

 

Dataset

Size (# of abstracts)

Training

True positive (TP)

3,536

 

True negative (TN)

1,959

 

Likely-positive (LP)

18,930

 

Unlabeled (U)

105,000

Test

Positive

338

 

Negative

339

Back to article page

BMC Bioinformatics

ISSN: 1471-2105

Contact us