Research on Named Entity Recognition of Electronic Medical Records Based on RoBERTa and Radical-Level Feature

Table 6

Comparison of different methods on CCKS2017.

Method	Precision	Recall	F1 score
Li et al. [33]	—	—	87.95
Ouyang et al. [34]	—	—	88.85
Qiu et al. [35]	90.63	92.02	91.32
Tang et al. [36]	88.60	94.25	91.34
Wang et al. [23]	90.83	91.64	91.24
Luo et al. [4]	—	—	91.75
Yin et al. [28]	92.30	93.28	92.79
Yin et al. [28]	92.27	93.73	93.00
Ours	94.14	92.39	93.26