A Chinese Named Entity Recognition Model of Maintenance Records for Power Primary Equipment Based on Progressive Multitype Feature Fusion

Table 4

Performance comparison of different methods on China People’s Daily corpus.

Methods	Metrics
	P (%)	R (%)	F1 (%)
Collobert et al. [35]	88.43	87.68	88.05
Chiu and Nichols [36]	91.94	90.06	90.99
Shen et al. [37]	90.45	89.72	90.08
Lample et al. [38]	91.46	90.18	90.81
The proposed method	97.51	95.93	96.54

Bold values show the best performance.