Abstract
Error type information has been widely used to improve the performance of grammatical error correction (GEC) models, whether for generating corrections, re-ranking them, or combining GEC models. Combining GEC models that have complementary strengths in correcting different error types is very effective in producing better corrections. However, system combination incurs a high computational cost due to the need to run inference on the base systems before running the combination method itself. Therefore, it would be more efficient to have a single model with multiple sub-networks that specialize in correcting different error types. In this paper, we propose a mixture-of-experts model, MoECE, for grammatical error correction. Our model successfully achieves the performance of T5-XL with three times fewer effective parameters. Additionally, our model produces interpretable corrections by also identifying the error type during inference.- Anthology ID:
- 2024.findings-emnlp.997
- Volume:
- Findings of the Association for Computational Linguistics: EMNLP 2024
- Month:
- November
- Year:
- 2024
- Address:
- Miami, Florida, USA
- Editors:
- Yaser Al-Onaizan, Mohit Bansal, Yun-Nung Chen
- Venue:
- Findings
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 17127–17138
- Language:
- URL:
- https://preview.aclanthology.org/add_missing_videos/2024.findings-emnlp.997/
- DOI:
- 10.18653/v1/2024.findings-emnlp.997
- Cite (ACL):
- Muhammad Reza Qorib, Alham Fikri Aji, and Hwee Tou Ng. 2024. Efficient and Interpretable Grammatical Error Correction with Mixture of Experts. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 17127–17138, Miami, Florida, USA. Association for Computational Linguistics.
- Cite (Informal):
- Efficient and Interpretable Grammatical Error Correction with Mixture of Experts (Qorib et al., Findings 2024)
- PDF:
- https://preview.aclanthology.org/add_missing_videos/2024.findings-emnlp.997.pdf