Document-level sentiment classification on Yelp 13
68.2AccuracyHN-ATT
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| HN-ATT2019.08 | 68.2 | — | |
| NSE-LSTMEncoder=NSE, Decoder=LSTM2016.07 | 67 | 0.47 | |
| MPAD-path2019.08 | 66.8 | — | |
| NSE-NSEEncoder=NSE, Decoder=NSE2016.07 | 66.6 | 0.48 | |
| MPAD-clique2019.08 | 66.6 | — | |
| MPAD-sentence-att2019.08 | 66.47 | — | |
| MPAD2019.08 | 66.16 | — | |
| LSTM-GRNN2016.07 | 65.1 | 0.5 | |
| LSTM-GRNN2019.08 | 65.1 | — | |
| CNN2019.08 | 64.89 | — | |
| Conv-GRNN2016.07 | 63.7 | 0.56 | |
| DAN2019.08 | 61.55 | — | |
| DiSAN2019.08 | 60.51 | — | |
| Classifier2016.07 | 59.8 | 0.68 | |
| CNN2016.07 | 59.7 | 0.76 | |
| PV2016.07 | 57.7 | 0.86 | |
| doc2vec2019.08 | 57.7 | — |