Irony Detection on SemEval Task3A 2018 (test)
74.6F1 (pos)BERTweet
Evaluation results
| Method | Links | ||
|---|---|---|---|
| BERTweetNormalization Strategy=soft2020.05 | 74.6 | 78.2 | |
| BERTweetNormalization Strategy=hard2020.05 | 74.3 | 78.2 | |
| RoBERTa-largeNormalization Strategy=soft2020.05 | 73.2 | 76.5 | |
| RoBERTa-largeNormalization Strategy=hard2020.05 | 71.9 | 75.1 | |
| RoBERTa-baseNormalization Strategy=hard2020.05 | 71.2 | 74 | |
| RoBERTa-baseNormalization Strategy=soft2020.05 | 71 | 74 | |
| XLM-R-largeNormalization Strategy=soft2020.05 | 70.8 | 74.2 | |
| Wu et al.2020.05 | 70.5 | 73.5 | |
| XLM-R-largeNormalization Strategy=hard2020.05 | 69.7 | 73.2 | |
| Baziotis et al.2020.05 | 67.2 | 73.2 | |
| XLM-R-baseNormalization Strategy=soft2020.05 | 66.6 | 70.8 | |
| XLM-R-baseNormalization Strategy=hard2020.05 | 66.2 | 70.8 |