Abstract
We present an error analysis of neural UPOS taggers to evaluate why using gold tags has such a large positive contribution to parsing performance while using predicted UPOS either harms performance or offers a negligible improvement. We also evaluate what neural dependency parsers implicitly learn about word types and how this relates to the errors taggers make, to explain the minimal impact using predicted tags has on parsers. We then mask UPOS tags based on errors made by taggers to tease away the contribution of UPOS tags that taggers succeed and fail to classify correctly and the impact of tagging errors.- Anthology ID:
- 2021.nodalida-main.31
- Volume:
- Proceedings of the 23rd Nordic Conference on Computational Linguistics (NoDaLiDa)
- Month:
- May 31--2 June
- Year:
- 2021
- Address:
- Reykjavik, Iceland (Online)
- Editors:
- Simon Dobnik, Lilja Øvrelid
- Venue:
- NoDaLiDa
- SIG:
- Publisher:
- Linköping University Electronic Press, Sweden
- Note:
- Pages:
- 309–314
- Language:
- URL:
- https://aclanthology.org/2021.nodalida-main.31
- DOI:
- Cite (ACL):
- Mark Anderson and Carlos Gómez-Rodríguez. 2021. What Taggers Fail to Learn, Parsers Need the Most. In Proceedings of the 23rd Nordic Conference on Computational Linguistics (NoDaLiDa), pages 309–314, Reykjavik, Iceland (Online). Linköping University Electronic Press, Sweden.
- Cite (Informal):
- What Taggers Fail to Learn, Parsers Need the Most (Anderson & Gómez-Rodríguez, NoDaLiDa 2021)
- PDF:
- https://preview.aclanthology.org/add_acl24_videos/2021.nodalida-main.31.pdf