Aspect Category Detection on SemEval Task 4 Subtask 3 2014 (test)
0.9357PrecisionBERT-pair-NLI-B
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| BERT-pair-NLI-BAuxiliary Sentence Type=NLI, Labeling Strategy=Binary2019.03 | 0.9357 | 0.9083 | 0.9218 | |
| BERT-pair-NLI-MAuxiliary Sentence Type=NLI, Labeling Strategy=Multi-label2019.03 | 0.9315 | 0.9024 | 0.9167 | |
| BERT-pair-QA-BAuxiliary Sentence Type=QA, Labeling Strategy=Binary2019.03 | 0.9304 | 0.8995 | 0.9147 | |
| BERT-pair-QA-MAuxiliary Sentence Type=QA, Labeling Strategy=Multi-label2019.03 | 0.9287 | 0.9024 | 0.9154 | |
| BERT-singleModel Type=Single sentence classification2019.03 | 0.9278 | 0.8907 | 0.9089 | |
| NRC-Canada2019.03 | 0.9104 | 0.8624 | 0.8858 | |
| XRCE2019.03 | 0.8323 | 0.8137 | 0.8229 |