Sarcasm Detection on Sarcasm Corpus Dialogues V2 (test)
83.6AccuracyOur Model
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Our ModelEncoder=RoBERTa, Evaluation Protocol=5-fold cross-validation2025.03 | 83.6 | 83.7 | 83.6 | |
| Our ModelEncoder=SBERT, Evaluation Protocol=5-fold cross-validation2025.03 | 82.1 | 82.5 | 82.4 | |
| Fuzzy-LogicEvaluation Protocol=5-fold cross-validation2025.03 | 81.8 | 80.3 | 81 | |
| BiGRUEvaluation Protocol=5-fold cross-validation2025.03 | 79.3 | 81.4 | 80.2 | |
| GRU-AttentionEvaluation Protocol=5-fold cross-validation2025.03 | 77.3 | 77.2 | 77.2 | |
| ELMo-BiLSTMArchitecture Variant=FULL, Evaluation Protocol=5-fold cross-validation2025.03 | 76.2 | 76 | 76 | |
| ELMo-BiLSTMEvaluation Protocol=5-fold cross-validation2025.03 | 74.6 | 74.7 | 74.7 | |
| MIARNEvaluation Protocol=5-fold cross-validation2025.03 | 74.2 | 72.9 | 72.7 | |
| SIARNEvaluation Protocol=5-fold cross-validation2025.03 | 71.9 | 71.8 | 71.8 | |
| CNN-LSTM-DNNEvaluation Protocol=5-fold cross-validation2025.03 | 67.3 | 66.7 | 65.7 | |
| GRNNEvaluation Protocol=5-fold cross-validation2025.03 | 64.4 | 61.8 | 61.2 |