Papers by Lucy H. Lin
Parsing with Multilingual BERT, a Small Corpus, and a Small Treebank (2020.findings-emnlp)
Copied to clipboard
| Challenge: | Pretrained multilingual contextual representations have shown great success, but their benefits do not apply equally to all language varieties. |
| Approach: | They propose to use additional language-specific pretraining and vocabulary augmentation to adapt multilingual models to low-resource settings. |
| Outcome: | The proposed methods significantly improve performance over baselines, especially in the lowest-resource cases, and demonstrate the importance of the relationship between such models’ pretraining data and target language varieties. |