Entity Linking on CWEB (test)
81.8Micro F1Yang et al.
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Yang et al.year=20182019.09 | 81.8 | — | |
| LUKEfine-tuning=none, ordering=confidence-order2019.09 | 78.9 | — | |
| Fang et al.2019.09 | 78.5 | — | |
| LUKEfine-tuning=none, ordering=natural-order2019.09 | 78.4 | — | |
| LUKEfine-tuning=none, ordering=local2019.09 | 78.4 | — | |
| LUKEfine-tuning=CoNLL dataset, ordering=confidence-order2019.09 | 78.3 | — | |
| LUKEfine-tuning=CoNLL dataset, ordering=natural-order2019.09 | 78.3 | — | |
| LUKEfine-tuning=CoNLL dataset, ordering=local2019.09 | 78.2 | — | |
| Deep-ED2019.09 | 77.9 | — | |
| Ganea and Hofmann2019.09 | 77.9 | — | |
| Barba et al.2019.09 | 77.7 | — | |
| Ment-Norm2019.09 | 77.5 | — | |
| Le and Titov2019.09 | 77.5 | — | |
| Cao et al.2019.09 | 77.3 | — | |
| WNED2019.09 | 77 | — | |
| Yang et al.year=20192019.09 | 75.6 | — | |
| ETHZ-Attn + DCA-RLBackbone=ETHZ-Attn, Learning protocol=Reinforcement Learning2019.09 | 75.59 | — | |
| ETHZ-Attn + DCA-SLBackbone=ETHZ-Attn, Learning protocol=Supervised Learning2019.09 | 73.47 | — | |
| Berkeley-CNN + DCA-RLBackbone=Berkeley-CNN, Learning protocol=Reinforcement Learning2019.09 | 72.66 | — | |
| Berkeley-CNN + DCA-SLBackbone=Berkeley-CNN, Learning protocol=Supervised Learning2019.09 | 71.01 | — | |
| ETHZ-AttnBackbone=ETHZ-Attn2019.09 | 70.07 | — | |
| Prior (p(e|m))2019.09 | 69.8 | — | |
| Berkeley-CNNBackbone=Berkeley-CNN2019.09 | 67.97 | — | |
| RI2019.09 | 67.5 | — | |
| AIDA2019.09 | 58.6 | — | |
| GLOW2019.09 | 56.2 | — | |
| BERT-Entity-Sim (local & global)2020.01 | — | 77.9 | |
| BERT+G&H's embeddings2020.01 | — | 75.7 | |
| Cheng and Roth (2013)2020.01 | — | 67.5 | |
| Ganea and Hofmann (2017)2020.01 | — | 77.9 | |
| Guo and Barbosa (2016)2020.01 | — | 77 | |
| Hoffart et al. (2011)2020.01 | — | 58.6 | |
| Le and Titov (2018)2020.01 | — | 77.5 | |
| Milne and Witten (2008)2020.01 | — | 64.1 | |
| Oracle type (FIGER)Type source=FIGER, Mode=Oracle2020.01 | — | 84.3 | |
| Oracle type (Ultra-fine)Type source=Ultra-fine, Mode=Oracle2020.01 | — | 85.8 | |
| Predict type (Ultra-fine)Type source=Ultra-fine, Mode=Predicted2020.01 | — | 77.8 | |
| Predict type (ZOE)Type source=ZOE, Mode=Predicted2020.01 | — | 77.6 | |
| prior p(e|m)2020.01 | — | 69.8 | |
| Ratinov et al. (2011)2020.01 | — | 56.2 |