Sentiment Analysis (Subtask A) on SemEval Task 4 2017 (test)
0.732Macro RecallBERTweet
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| BERTweetNormalization=soft2020.05 | 0.732 | 72.8 | 71.7 | |
| BERTweetNormalization=hard2020.05 | 0.728 | 72.5 | 72 | |
| RoBERTa_largeNormalization=soft2020.05 | 0.725 | 72 | 70.7 | |
| RoBERTa_largeNormalization=hard2020.05 | 0.722 | 71.8 | 71.3 | |
| RoBERTa_baseNormalization=hard2020.05 | 0.718 | 71.2 | 70.9 | |
| XLM-R_largeNormalization=soft2020.05 | 0.717 | 71.1 | 70.7 | |
| XLM-R_largeNormalization=hard2020.05 | 0.717 | 70.9 | 70.6 | |
| RoBERTa_baseNormalization=soft2020.05 | 0.716 | 71.2 | 71.6 | |
| XLM-R_baseNormalization=soft2020.05 | 0.703 | 69.4 | 69.3 | |
| XLM-R_baseNormalization=hard2020.05 | 0.703 | 69.6 | 69.7 | |
| BB_twtrRank=1/382017.04 | 0.681 | — | — | |
| Next best submission2017.04 | 0.681 | — | — | |
| Cliche2020.05 | 0.681 | 68.5 | 65.8 | |
| Baziotis et al.2020.05 | 0.681 | 67.7 | 65.1 |