Skip to main content

Advertisement

Table 2 Performance values in terms of the precision (%), recall (%) and F1-score (%) for the state-of-the-art methods and the proposed model DTranNER

From: DTranNER: biomedical named entity recognition with deep learning-based label-label transition model

CorpusBC2GMBC4CHEMDBC5CDR-ChemicalBC5CDR-DiseaseNCBI-Disease
 PRF1PRF1PRF1PRF1PRF1
Att-BiLSTM-CRF (2017)---92.2990.0191.1493.4991.6892.57------
D3NER (2018)------93.7392.5693.1483.9885.4084.6885.0383.8084.41
Collabonet (2018)80.4978.9979.7390.7887.0188.8594.2692.3893.3185.6182.6184.0885.4887.2786.36
Wang et al. (2018)82.1079.4280.7491.3087.5389.3793.5692.4893.0384.1485.7684.9585.8686.4286.14
BioBERT (2019)85.1683.6584.4092.2390.6191.4193.2793.6193.4485.8687.2786.5689.0489.6989.36
DTranNER84.2184.8484.5691.9492.0491.9994.2894.0494.1686.7587.7087.2288.2189.0488.62
  1. Note: The highest performance in each corpus is highlighted in Bold. We quoted the published scores for the other models. For Wang et al. [11], we conducted additional experiments to obtain the performance scores for two corpora (i.e., BC5CDR-Chemical and BC5CDR-Disease) using the software on their open source repository [45]