Named Entity Recognition on GENIA (test)
81.39F1 ScoreW2NER
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| W2NERMethod Category=Proposed Method2021.12 | 81.39 | 83.1 | 79.76 | — | |
| Shen et al.Method Category=Span-based2021.12 | 80.54 | 80.19 | 80.89 | — | |
| Yu et al.Method Category=Span-based2021.12 | 80.5 | 81.8 | 79.3 | — | |
| Yan et al.Method Category=Seq2Seq2021.12 | 79.23 | 78.87 | 79.6 | — | |
| Wang et al.Method Category=Span-based2021.12 | 79.19 | 79.45 | 78.94 | — | |
| Seq2Seq, ELMo, BERT, FlairReference=Straková et al. (2019)2021.06 | 78.3 | — | — | — | |
| Straková et al.Method Category=Seq2Seq2021.12 | 78.2 | — | — | — | |
| DYGIE++Backbone=BERT-Base2019.09 | 77.9 | — | — | — | |
| Span-based, BERT, CorefReference=Wadden et al. (2019)2021.06 | 77.9 | — | — | — | |
| Span-based, BERT (Our Model)Embeddings=BERT, Architecture=Span-based2021.06 | 77.8 | — | — | — | |
| Span-based model (- Dep guided GCN)Ablation=w/o Dependency-guided GCN2021.06 | 77.4 | — | — | — | |
| Span-based model (- Overlap Relation)Ablation=w/o Overlap Relation2021.06 | 77.4 | — | — | — | |
| DYGIEBackbone=ELMo2019.09 | 76.2 | — | — | — | |
| Span-based, ELMo, CorefReference=Luan et al. (2019)2021.06 | 76.2 | — | — | — | |
| BiFlaG2020.05 | 76 | 77.4 | 74.6 | — | |
| Segm. hypPOS tags=true2020.05 | 75.1 | 77 | 73.3 | — | |
| Hypergraph, RNNReference=Wang and Lu (2018)2021.06 | 75.1 | — | — | — | |
| Wang and LuMethod Category=Hypergraph-based2021.12 | 75.1 | 77 | 73.3 | — | |
| Anchor-RegionPOS tags=true2020.05 | 74.8 | 75.8 | 73.9 | — | |
| Anchor-region detection, RNN, CNNReference=Lin et al. (2019a)2021.06 | 74.8 | — | — | — | |
| layered-CRF2020.05 | 74.7 | 78.5 | 71.3 | — | |
| Boundary-aware2020.05 | 74.7 | 75.9 | 73.6 | — | |
| Dynamically stacking, RNNReference=Ju et al. (2018)2021.06 | 74.7 | — | — | — | |
| Boundary detection, RNNReference=Zheng et al. (2019)2021.06 | 74.7 | — | — | — | |
| Ju et al.Method Category=Sequence Labeling2021.12 | 74.7 | 78.5 | 71.3 | — | |
| Transition-based parsing, RNNReference=Wang et al. (2018)2021.06 | 73.9 | — | — | — | |
| Hypergraph, RNNReference=Katiyar and Cardie (2018)2021.06 | 73.8 | — | — | — | |
| LSTM. hyp2020.05 | 73.6 | 79.8 | 68.2 | — | |
| HypergraphReference=Muis and Lu (2017)2021.06 | 70.8 | — | — | — | |
| Exhaustive2020.05 | 70.7 | 73.3 | 68.3 | — | |
| Constituency parsingReference=Finkel and Manning (2009)2021.06 | 70.3 | — | — | — | |
| HypergraphReference=Lu and Roth (2015)2021.06 | 70.3 | — | — | — | |
| LSTM-CRF2020.05 | 69.5 | 75.2 | 64.6 | — | |
| Multi-CRF2020.05 | 68.8 | 73.1 | 64.9 | — | |
| Meta(SpanPred U SEQ)combiner=Meta2023.05 | — | 78.3 | 77.4 | 79.2 | |
| SEQ2023.05 | — | 76.1 | 79.8 | 72.7 | |
| SeqCRF2023.05 | — | 75.7 | 79.7 | 72.1 | |
| SOTA (Shen et al., 2022)2023.05 | — | 81.7 | — | — | |
| SpanPred2023.05 | — | 77.1 | 77 | 77.1 | |
| SpanPred U SEQcombiner=Union2023.05 | — | 77.2 | 73.5 | 81.4 | |
| SpanPred U SEQ U SeqCRFcombiner=Union2023.05 | — | 76.4 | 71.3 | 82.3 | |
| SpanPred x SEQcombiner=MajVote2023.05 | — | 75.8 | 85 | 68.5 | |
| SpanPred x SEQ x SeqCRFcombiner=MajVote2023.05 | — | 77.1 | 81.9 | 72.9 |