Skip to main content

Table 4 The number of bio-entity type errors, the total number of errors, and the ratio of bio-entity errors to the total numbers of errors for each model prediction

From: CollaboNet: collaboration of deep neural networks for biomedical named entity recognition

  Our STM CollaboNet  
Dataset Bio Entity Total Ratio of Bio Entity Bio Entity Total Ratio of Bio Entity Difference
NCBI-disease 54 167 32.3% 38 131 29.0% -3.3%
JNLPBA 749 1520 49.3% 227 1437 15.8% -33.5%
BC5CDR-chem 142 503 28.2% 122 505 24.2% -4.1%
BC5CDR-disease 199 867 23.0% 131 728 18.0% -5.0%
BC2GM 189 1277 14.8% 218 1165 18.7% 3.9%
  1. Negative values at the difference tab indicate that CollaboNet reduced the number of false positives, especially false biomedical entities