Joint Entity and Relation Extraction on ADE
0.912Entity F1 ScoreTANL
Evaluation results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| TANLTraining Setting=multi-task2021.01 | 0.912 | — | — | — | — | 0.838 | — | — | — | |
| TANL (multitask)mode=multitask2022.05 | 0.912 | — | — | — | — | 0.838 | — | — | — | |
| TANL2022.05 | 0.912 | — | — | — | — | 0.838 | — | — | — | |
| DEEPSTRUCTEvaluation protocol=w/ finetune2022.05 | 0.911 | — | — | — | — | 0.838 | — | — | — | |
| DEEPSTRUCTmode=w/ finetune2022.05 | 0.911 | — | — | — | — | 0.838 | — | — | — | |
| DEEPSTRUCTprotocol=w/ finetune2022.05 | 0.911 | — | — | — | — | 0.838 | — | — | — | |
| DEEPSTRUCTEvaluation protocol=multi-task2022.05 | 0.905 | — | — | — | — | 0.836 | — | — | — | |
| DEEPSTRUCTmode=multi-task2022.05 | 0.905 | — | — | — | — | 0.836 | — | — | — | |
| DEEPSTRUCTprotocol=multi-task2022.05 | 0.905 | — | — | — | — | 0.836 | — | — | — | |
| TANLTraining Setting=standard2021.01 | 0.902 | — | — | — | — | 0.806 | — | — | — | |
| TANL2022.05 | 0.902 | — | — | — | — | 0.806 | — | — | — | |
| TANLTraining Setting=multi-dataset2021.01 | 0.9 | — | — | — | — | 0.8 | — | — | — | |
| SpERTTraining Setting=standard2021.01 | 0.893 | — | — | — | — | 0.788 | — | — | — | |
| SpERT2022.05 | 0.893 | — | — | — | — | 0.788 | — | — | — | |
| Eberts and Ulges, 20202022.05 | 0.893 | — | — | — | — | 0.788 | — | — | — | |
| multi-headPre-calculated Features=false, Evaluation Protocol=strict2018.04 | 0.864 | 0.8472 | 0.8816 | 0.721 | 0.7724 | 0.7458 | 0.8049 | — | — | |
| Li et al. (2017)Pre-calculated Features=false, Evaluation Protocol=strict2018.04 | 0.846 | 0.827 | 0.867 | 0.675 | 0.758 | 0.714 | 0.78 | — | — | |
| baseline + ATExternal NLP Tools=false, Evaluation Protocol=Standard2018.08 | 0.8361 | — | — | — | — | 0.6195 | 0.7278 | — | — | |
| baselineExternal NLP Tools=false, Evaluation Protocol=Standard2018.08 | 0.8304 | — | — | — | — | 0.6104 | 0.7204 | — | — | |
| Miwa and Sasaki (2014)External NLP Tools=true, Evaluation Protocol=Standard2018.08 | 0.807 | — | — | — | — | 0.61 | 0.7085 | — | — | |
| Li et al. (2016)Pre-calculated Features=false, Evaluation Protocol=strict2018.04 | 0.795 | 0.795 | 0.796 | 0.64 | 0.629 | 0.634 | 0.7145 | — | — | |
| DEEPSTRUCTEvaluation protocol=zero-shot2022.05 | 0.607 | — | — | — | — | 0.106 | — | — | — | |
| DEEPSTRUCTmode=zero-shot2022.05 | 0.607 | — | — | — | — | 0.106 | — | — | — | |
| DEEPSTRUCTprotocol=zero-shot2022.05 | 0.607 | — | — | — | — | 0.106 | — | — | — | |
| GPT-3 175Bmode=zero-shot2022.05 | 0.058 | — | — | — | — | 0.013 | — | — | — | |
| GPT-3 175Bprotocol=zero-shot2022.05 | 0.058 | — | — | — | — | 0.013 | — | — | — | |
| BiTT-MedEncoder=BERT-BASE2020.08 | — | — | — | — | — | — | 0.821 | 0.831 | 0.813 | |
| Multi-headEncoder=LSTM2020.08 | — | — | — | — | — | — | 0.745 | 0.721 | 0.772 | |
| Multi-head + ATEncoder=LSTM2020.08 | — | — | — | — | — | — | 0.755 | — | — | |
| Neural JointEncoder=LSTM2020.08 | — | — | — | — | — | — | 0.634 | 0.64 | 0.629 | |
| PFNEncoder=BERT-BASE2020.08 | — | — | — | — | — | — | 0.8 | — | — | |
| Rel-MetricEncoder=LSTM + CNN2020.08 | — | — | — | — | — | — | 0.773 | 0.774 | 0.773 | |
| Table-SequenceEncoder=ALBERT-XXLARGE-V12020.08 | — | — | — | — | — | — | 0.801 | — | — |