Entity Disambiguation on AQUAINT (test)
93.5Micro F1LUKE
Evaluation Results
| Method | Links | |
|---|---|---|
| LUKEfine-tuning=none, ordering=confidence-order2019.09 | 93.5 | |
| LUKEfine-tuning=none, ordering=natural-order2019.09 | 92.9 | |
| LUKEfine-tuning=none, ordering=local2019.09 | 91.9 | |
| Barba et al.2019.09 | 91.6 | |
| LUKEfine-tuning=CoNLL dataset, ordering=confidence-order2019.09 | 91.5 | |
| LUKEfine-tuning=CoNLL dataset, ordering=natural-order2019.09 | 90.9 | |
| LUKEfine-tuning=CoNLL dataset, ordering=local2019.09 | 90.8 | |
| Yang et al.year=20182019.09 | 89.9 | |
| Cao et al.2019.09 | 89.9 | |
| Ganea and Hofmann2019.09 | 88.5 | |
| Le and Titov2019.09 | 88.3 | |
| Yang et al.year=20192019.09 | 88.3 | |
| Fang et al.2019.09 | 87.5 | |
| CHOLANTraining Dataset=CONLL-AIDA, Candidate Generation=Falcon2021.01 | 0.768 | |
| CHOLAN-Wiki+ DCATraining Dataset=CONLL-AIDA, Candidate Generation=DCA2021.01 | 0.759 | |
| CHOLAN-Wiki+ FCTraining Dataset=CONLL-AIDA, Candidate Generation=Falcon2021.01 | 0.7 | |
| Mendes et al. 2011Training Dataset=CONLL-AIDA2021.01 | 0.452 | |
| Kolitsas et al. 2018Training Dataset=CONLL-AIDA2021.01 | 0.404 | |
| Steinmetz and Sack 2013Training Dataset=CONLL-AIDA2021.01 | 0.359 | |
| Moro et al. 2014Training Dataset=CONLL-AIDA2021.01 | 0.358 |