Abstract
We introduce the Korean-Learner-Morpheme (KLM) corpus, a manually annotated dataset consisting of 129,784 morphemes from second language (L2) learners of Korean, featuring morpheme tokenization and part-of-speech (POS) tagging. We evaluate the performance of four Korean morphological analyzers in tokenization and POS tagging on the L2- Korean corpus. Results highlight the analyzers’ reduced performance on L2 data, indicating the limitation of advanced deep-learning models when dealing with L2-Korean corpora. We further show that fine-tuning one of the models with the KLM corpus improves its accuracy of tokenization and POS tagging on L2-Korean dataset.- Anthology ID:
- 2023.bea-1.6
- Volume:
- Proceedings of the 18th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2023)
- Month:
- July
- Year:
- 2023
- Address:
- Toronto, Canada
- Editors:
- Ekaterina Kochmar, Jill Burstein, Andrea Horbach, Ronja Laarmann-Quante, Nitin Madnani, Anaïs Tack, Victoria Yaneva, Zheng Yuan, Torsten Zesch
- Venue:
- BEA
- SIG:
- SIGEDU
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 72–82
- Language:
- URL:
- https://aclanthology.org/2023.bea-1.6
- DOI:
- 10.18653/v1/2023.bea-1.6
- Cite (ACL):
- Hakyung Sung and Gyu-Ho Shin. 2023. Towards L2-friendly pipelines for learner corpora: A case of written production by L2-Korean learners. In Proceedings of the 18th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2023), pages 72–82, Toronto, Canada. Association for Computational Linguistics.
- Cite (Informal):
- Towards L2-friendly pipelines for learner corpora: A case of written production by L2-Korean learners (Sung & Shin, BEA 2023)
- PDF:
- https://preview.aclanthology.org/landing_page/2023.bea-1.6.pdf