Fine-grained Entity Typing on FIGER (test)
86.7Macro F1LITE
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| LITETraining/Evaluation Protocol=NLI+task-specific training2022.02 | 86.7 | — | 83.3 | |
| SEPREM2022.02 | 86.1 | — | 82.1 | |
| K-ADAPTER (F+L)Knowledge Type=Factual + Linguistic2020.02 | 84.87 | 61.81 | 80.54 | |
| K-ADAPTER (F)Knowledge Type=Factual2020.02 | 84.52 | 59.5 | 80.42 | |
| ROBERTa + multitask2020.02 | 84.45 | 59.86 | 78.84 | |
| K-ADAPTER (L)Knowledge Type=Linguistic2020.02 | 83.61 | 61.1 | 79.18 | |
| DSAM2022.02 | 83.3 | — | 81.5 | |
| Lin and Ji (2019)large-scale augmented data=true2021.01 | 83 | — | 79.8 | |
| Hierarchy-Typing2022.02 | 83 | — | 79.8 | |
| Chen et al. (2020) (exclusive)2021.01 | 82.6 | — | 80.8 | |
| K-ADAPTER (w/o knowledge)Knowledge Type=None2020.02 | 82.56 | 56.93 | 77.9 | |
| ROBERTa2020.02 | 82.43 | 56.31 | 77.83 | |
| WKLM2020.02 | 81.99 | 60.21 | 77 | |
| Vector2021.01 | 81.6 | — | 77 | |
| Chen et al. (2020) (undefined)2021.01 | 80.5 | — | 78.1 | |
| LITETraining/Evaluation Protocol=pre-trained on NLI+UFET2022.02 | 80.1 | — | 74.7 | |
| Box2021.01 | 79.4 | — | 75 | |
| Box4Types2022.02 | 79.4 | — | 75 | |
| Zhang et al. (2018)2021.01 | 78.7 | — | 75.5 | |
| ERNIE2020.02 | 75.61 | 57.19 | 73.39 | |
| BERT-base2020.02 | 75.16 | 52.04 | 71.63 | |
| NFGEC2020.02 | 75.15 | 55.6 | 71.73 |