Aspect-level sentiment analysis on Rest16
93.02AccuracyHGCN+BERT
Evaluation results
| Method | Links | |
|---|---|---|
| HGCN+BERTBackbone=BERT, Evaluation Protocol=Best performance2022.04 | 93.02 | |
| HGCN+BERTBackbone=BERT, Evaluation Protocol=Average performance2022.04 | 91.72 | |
| DualGCN+BERTBackbone=BERT2022.04 | 91.43 | |
| RGAT+BERTBackbone=BERT2022.04 | 91.39 | |
| DGEDT+BERTBackbone=BERT2022.04 | 90.94 | |
| BERT+FinetuneBackbone=BERT2022.04 | 90.24 | |
| HGCNBackbone=GloVe + BiLSTM, Evaluation Protocol=Best performance2022.04 | 89.84 | |
| InstructABSAvariant=22023.02 | 89.43 | |
| InstructABSAvariant=12023.02 | 89.1 | |
| HGCNBackbone=GloVe + BiLSTM, Evaluation Protocol=Average performance2022.04 | 88.92 | |
| InterGCNBackbone=GloVe + BiLSTM2022.04 | 88.52 | |
| BiGCNBackbone=GloVe + BiLSTM2022.04 | 88.47 | |
| RGATBackbone=GloVe + BiLSTM2022.04 | 88.21 | |
| DualGCNBackbone=GloVe + BiLSTM2022.04 | 88.19 | |
| ASGCNBackbone=GloVe + BiLSTM2022.04 | 88.02 | |
| MCRF-SABackbone=GloVe + BiLSTM2022.04 | 87.76 | |
| AOABackbone=GloVe + BiLSTM2022.04 | 87.5 | |
| DGEDTBackbone=GloVe + BiLSTM2022.04 | 86.52 | |
| SA-LSTMBackbone=GloVe + BiLSTM2022.04 | 85.96 | |
| MGANBackbone=GloVe + BiLSTM2022.04 | 85.29 | |
| ATAE-LSTMBackbone=GloVe + BiLSTM2022.04 | 84.36 | |
| CDTBackbone=GloVe + BiLSTM2022.04 | 83.88 |