Readability Classification on WeeBit (test)
90.5AccuracyBART-RF-T1
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| BART-RF-T1Backbone=BART, Classifier=Random Forest, Feature Set=T1, Mode=Hybrid2021.09 | 90.5 | 0.905 | 0.904 | 90.5 | 0.968 | |
| RoBERTa-RF-T1Backbone=RoBERTa, Classifier=Random Forest, Feature Set=T1, Mode=Hybrid2021.09 | 90.2 | 0.903 | 0.903 | 90.2 | 0.971 | |
| BERT-GB-T1Backbone=BERT, Classifier=XGBoost, Feature Set=T1, Mode=Hybrid2021.09 | 89.5 | 0.897 | 0.897 | 89.5 | 0.969 | |
| XLNet-RF-P3Backbone=XLNet, Classifier=Random Forest, Feature Set=P3, Mode=Hybrid2021.09 | 89.2 | 0.893 | 0.892 | 89.2 | 0.966 | |
| BERT2019.07 | 85.73 | 86.58 | 85.73 | 85.81 | 0.9527 | |
| Mar-21 (BERT)Model=BERT2021.09 | 85.7 | 0.857 | 0.858 | 86.6 | 0.953 | |
| Filighera et al. (2019)2019.07 | 81.3 | — | — | — | — | |
| Fili-19 (LSTM)Model=LSTM2021.09 | 81.3 | — | — | — | — | |
| Xia et al. (2016)2019.07 | 80.3 | — | — | — | — | |
| Xia-16 (SVM)Model=SVM2021.09 | 80.3 | — | — | 81.3 | — | |
| BiLSTM2019.07 | 77.43 | 78.02 | 77.43 | 77.5 | 0.906 | |
| HAN2019.07 | 75.2 | 75.34 | 75.2 | 75.2 | 0.886 | |
| Mar-21 (HAN)Model=HAN2021.09 | 75.2 | 0.752 | 0.752 | 75.3 | 0.886 | |
| SVM-BF (Deutsch et al., 2020)2019.07 | — | — | — | 83.81 | — |