Table 2 Best positive effects
From: A neural network multi-task learning approach to biomedical named entity recognition
Dataset | STM | Best MO-MTM | Best dataset |
|---|---|---|---|
AnatEM | 81.55 | 81.68 | NCBI-Disease |
BC2GM | 72.63 | 72.21 | Ex-PTM |
BC4CHEMD | 82.95 | 80.31 | BioNLP13GE |
BC5CDR | 83.66 | 83.77 | BioNLP11EPI |
BioNLP09 | 83.90 | 84.16 | BioNLP13GE |
BioNLP11EPI | 77.72 | 78.10 | BioNLP09 |
BioNLP11ID | 81.50 | 82.26* | BioNLP13GE |
BioNLP13CG | 76.74 | 77.33* | BioNLP13PC |
BioNLP13GE | 73.28 | 76.09* | BioNLP11EPI |
BioNLP13PC | 80.61 | 80.94 | Ex-PTM |
CRAFT | 79.55 | 78.48 | BioNLP13GE |
Ex-PTM | 68.56 | 73.58* | BioNLP11EPI |
JNLPBA | 69.60 | 68.92 | BioNLP13GE |
Linnaeus | 83.98 | 83.63 | NCBI-Disease |
NCBI-Disease | 80.26 | 80.74 | Ex-PTM |
Average | 78.43 | 78.81 | N/A |