Aspect-level Sentiment Classification on SemEval Restaurant 2014 (test)
91.43AccuracyLSAE-DeBERTa
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| LSAE-DeBERTaBackbone=DeBERTa2021.10 | 91.43 | — | — | — | 86.85 | |
| LSAE-X-DeBERTaBackbone=X-DeBERTa2021.10 | 90.98 | — | — | — | 87.02 | |
| LSAT-X-DeBERTaBackbone=X-DeBERTa2021.10 | 90.86 | — | — | — | 86.26 | |
| LSAS-X-DeBERTaBackbone=X-DeBERTa2021.10 | 90.33 | — | — | — | 85.55 | |
| LSAP-X-DeBERTaBackbone=X-DeBERTa2021.10 | 90.27 | — | — | — | 85.51 | |
| LSAP-DeBERTaBackbone=DeBERTa2021.10 | 89.91 | — | — | — | 84.9 | |
| LSAT-DeBERTaBackbone=DeBERTa2021.10 | 89.91 | — | — | — | 85.05 | |
| LSAS-DeBERTaBackbone=DeBERTa2021.10 | 89.64 | — | — | — | 84.53 | |
| LSAE-RoBERTaBackbone=RoBERTa2021.10 | 89.11 | — | — | — | 83.98 | |
| DeBERTaBackbone=DeBERTa2021.10 | 88.66 | — | — | — | 83.06 | |
| LSAS-RoBERTaBackbone=RoBERTa2021.10 | 88.48 | — | — | — | 83.81 | |
| LSAT-RoBERTaBackbone=RoBERTa2021.10 | 88.3 | — | — | — | 83.09 | |
| SARL-RoBERTaBackbone=RoBERTa2021.10 | 88.21 | — | — | — | 82.44 | |
| LSAP-RoBERTaBackbone=RoBERTa2021.10 | 88.04 | — | — | — | 82.96 | |
| RoBERTa + SKEPModel Size=large2020.05 | 88.01 | — | — | — | — | |
| RoBERTa + SKEPModel Size=base2020.05 | 87.92 | — | — | — | — | |
| BERT-ADA RestTrain Dataset=Lapt. + Rest., Train Type=Joint2019.08 | 87.89 | — | — | 81.05 | — | |
| Previous SOTA2020.05 | 87.89 | — | — | — | — | |
| RoBERTaBackbone=RoBERTa2021.10 | 87.77 | — | — | — | 82.1 | |
| BERT-ADA JointTrain Dataset=Lapt. + Rest., Train Type=Joint2019.08 | 87.69 | — | — | 81.2 | — | |
| RGAT-RoBERTaBackbone=RoBERTa2021.10 | 87.52 | — | — | — | 81.29 | |
| LSAS-BERTBackbone=BERT2021.10 | 87.41 | — | — | — | 81.52 | |
| LSAE-BERTBackbone=BERT2021.10 | 87.41 | — | — | — | 81.52 | |
| PWCN-RoBERTaBackbone=RoBERTa2021.10 | 87.35 | — | — | — | 80.85 | |
| LSAT-BERTBackbone=BERT2021.10 | 87.32 | — | — | — | 81.86 | |
| SSEGCN-BERTBackbone=BERT2021.10 | 87.31 | — | — | — | 81.09 | |
| LSAP-BERTBackbone=BERT2021.10 | 87.23 | — | — | — | 81.06 | |
| BERT-ADA RestTrain Dataset=Restaurants, Train Type=In-domain2019.08 | 87.14 | — | — | 80.05 | — | |
| DualGCN-BERTBackbone=BERT2021.10 | 87.13 | — | — | — | 81.16 | |
| TF-BERTBackbone=BERT2021.10 | 87.09 | — | — | — | 81.15 | |
| ASGCN-RoBERTaBackbone=RoBERTa2021.10 | 86.87 | — | — | — | 80.59 | |
| SARL-DeBERTaBackbone=DeBERTa2021.10 | 86.69 | — | — | — | 78.91 | |
| BERT-ADA JointTrain Dataset=Restaurants, Train Type=In-domain2019.08 | 86.35 | — | — | 78.89 | — | |
| DGEDT-BERTBackbone=BERT2021.10 | 86.3 | — | — | — | 80 | |
| BERT-ADA LaptTrain Dataset=Lapt. + Rest., Train Type=Joint2019.08 | 86.22 | — | — | 79.79 | — | |
| dotGCN-BERTBackbone=BERT2021.10 | 86.16 | — | — | — | 80.49 | |
| TGCN-BERTBackbone=BERT2021.10 | 86.16 | — | — | — | 79.95 | |
| XLNet-baseTrain Dataset=Lapt. + Rest., Train Type=Joint2019.08 | 86.15 | — | — | 78.93 | — | |
| BAT2020.01 | 86.03 | — | — | 79.24 | — | |
| BERT-PTselection=best2020.01 | 85.92 | — | — | 79.12 | — | |
| RoBERTaModel Size=large2020.05 | 85.88 | — | — | — | — | |
| XLNet-baseTrain Dataset=Restaurants, Train Type=In-domain2019.08 | 85.84 | — | — | 78.35 | — | |
| BERT-ADA LaptTrain Dataset=Restaurants, Train Type=In-domain2019.08 | 85.51 | — | — | 78.09 | — | |
| BERT-baseTrain Dataset=Lapt. + Rest., Train Type=Joint2019.08 | 85.03 | — | — | 77.35 | — | |
| BERT-PTTrain Dataset=Restaurants, Train Type=In-domain2019.08 | 84.95 | — | — | 76.96 | — | |
| BERT-PT2020.01 | 84.95 | — | — | 76.96 | — | |
| RoBERTaModel Size=base2020.05 | 84.93 | — | — | — | — | |
| BERT-baseTrain Dataset=Restaurants, Train Type=In-domain2019.08 | 84.92 | — | — | 76.93 | — | |
| BERT-SPCTrain Dataset=Restaurants, Train Type=In-domain2019.08 | 84.46 | — | — | 76.98 | — | |
| BERT-ADA RestTrain Dataset=Laptops, Train Type=Cross-domain2019.08 | 83.68 | — | — | 72.91 | — | |
| SDGCN-BERTTrain Dataset=Restaurants, Train Type=In-domain2019.08 | 83.57 | — | — | 76.47 | — | |
| SDGCN-BERTBackbone=BERT2021.10 | 83.57 | — | — | — | 76.47 | |
| SK-GCN-BERTBackbone=BERT2021.10 | 83.48 | — | — | — | 75.19 | |
| AEN-BERTTrain Dataset=Restaurants, Train Type=In-domain2019.08 | 83.12 | — | — | 73.76 | — | |
| XLNet-baseTrain Dataset=Laptops, Train Type=Cross-domain2019.08 | 82.41 | — | — | 72.98 | — | |
| BERT-ADA JointTrain Dataset=Laptops, Train Type=Cross-domain2019.08 | 82.23 | — | — | 73.03 | — | |
| BERT2020.01 | 81.54 | — | — | 71.94 | — | |
| MGAN2020.01 | 81.49 | — | — | 71.48 | — | |
| AOA-LSTMruns=102018.04 | 81.2 | 79.7 | 0.008 | — | — | |
| BERT-ADA LaptTrain Dataset=Laptops, Train Type=Cross-domain2019.08 | 80.68 | — | — | 72.93 | — | |
| BERT-baseTrain Dataset=Laptops, Train Type=Cross-domain2019.08 | 80.07 | — | — | 69.93 | — | |
| IAN2018.04 | 78.6 | — | — | — | — | |
| ATAE-LSTM2018.04 | 77.2 | — | — | — | — | |
| AT-LSTM2018.04 | 76.2 | — | — | — | — | |
| TD-LSTM2018.04 | 75.6 | — | — | — | — | |
| LSTM2018.04 | 74.3 | — | — | — | — | |
| Majority2018.04 | 53.5 | — | — | — | — | |
| BARTABSAModel=BARTABSA2024.05 | — | — | — | — | 75.56 | |
| Dual-MRCModel=Dual-MRC2024.05 | — | — | — | — | 82.04 | |
| InstructABSA2Model=InstructABSA2, Status=Reproduced2024.05 | — | — | — | — | 86.7 | |
| LSAT-XModel=LSAT-X2024.05 | — | — | — | — | 86.26 | |
| PFInstruct-NERModel=PFInstruct-NER, Prefix=Named Entity Recognition2024.05 | — | — | — | — | 86.66 | |
| PFInstruct-NoiseModel=PFInstruct-Noise, Prefix=Noise2024.05 | — | — | — | — | 86.88 | |
| PFInstruct-REModel=PFInstruct-RE, Prefix=Relation Extraction2024.05 | — | — | — | — | 86.68 |