Table 7 Performance comparison of different methods on multiple datasets. Bold values indicate the best results.
From: A multi-level feature enhancement framework for named entity recognition in power system texts
CoNLL-2003 | Resume | China People’s Daily Corpus | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
Model | P(%) | R(%) | F1(%) | Model | P(%) | R(%) | F1(%) | Model | P(%) | R(%) | F1(%) |
Zhang et al.45 | 94.71 | 92.42 | 93.64 | Guo et al.46 | 94.9 | 94.56 | 94.62 | Zhang et al.47 | 93.23 | 92.42 | 92.82 |
Shah et al.48 | 93.13 | 95.67 | 95.3 | Wang et al.49 | – | – | 96.53 | Zhang et al.45 | 94.71 | 92.42 | 93.64 |
Chen et al.50 | – | – | 93.09 | Mai et al.51 | 96.91 | 96.26 | 96.58 | Wang et al.52 | – | – | 95.32 |
Fei et al.53 | 92.96 | 93.85 | 93.4 | Chen et al.54 | 96.14 | 96.52 | 96.33 | Liu et al.55 | 92.17 | 90.63 | 91.4 |
Chang et al.56 | – | – | 93.46 | Zhang et al.57 | 96.09 | 96.44 | 96.26 | Lv et al.58 | 97.12 | 96.11 | 96.61 |
Yu et al.59 | – | – | 93.42 | Chen et al.60 | 96.54 | 96.41 | 96.58 | Wang et al.61 | – | – | 95.62 |
Gan et al.62 | 93.84 | 93.6 | 93.72 | Pan et al.63 | – | – | 96.35 | Ke et al.64 | 95.93 | 96.45 | 96.19 |
Ours | 94.65 | 94.44 | 94.55 | Ours | 96.54 | 97.25 | 96.89 | Ours | 97.45 | 96.93 | 97.19 |
