Abstract
Grammatical error correction can be viewed as a low-resource sequence-to-sequence task, because publicly available parallel corpora are limited. To tackle this challenge, we first generate erroneous versions of large unannotated corpora using a realistic noising function. The resulting parallel corpora are sub-sequently used to pre-train Transformer models. Then, by sequentially applying transfer learning, we adapt these models to the domain and style of the test set. Combined with a context-aware neural spellchecker, our system achieves competitive results in both restricted and low resource tracks in ACL 2019 BEAShared Task. We release all of our code and materials for reproducibility.- Anthology ID:
- W19-4423
- Volume:
- Proceedings of the Fourteenth Workshop on Innovative Use of NLP for Building Educational Applications
- Month:
- August
- Year:
- 2019
- Address:
- Florence, Italy
- Editors:
- Helen Yannakoudakis, Ekaterina Kochmar, Claudia Leacock, Nitin Madnani, Ildikó Pilán, Torsten Zesch
- Venue:
- BEA
- SIG:
- SIGEDU
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 213–227
- Language:
- URL:
- https://aclanthology.org/W19-4423
- DOI:
- 10.18653/v1/W19-4423
- Cite (ACL):
- Yo Joong Choe, Jiyeon Ham, Kyubyong Park, and Yeoil Yoon. 2019. A Neural Grammatical Error Correction System Built On Better Pre-training and Sequential Transfer Learning. In Proceedings of the Fourteenth Workshop on Innovative Use of NLP for Building Educational Applications, pages 213–227, Florence, Italy. Association for Computational Linguistics.
- Cite (Informal):
- A Neural Grammatical Error Correction System Built On Better Pre-training and Sequential Transfer Learning (Choe et al., BEA 2019)
- PDF:
- https://preview.aclanthology.org/nschneid-patch-4/W19-4423.pdf
- Code
- kakaobrain/helo_word + additional community code
- Data
- WI-LOCNESS, WikiText-103, WikiText-2