Abstract
Cross-lingual and multilingual methods have been widely suggested as options for dependency parsing of low-resource languages; however, these typically require the use of annotated data in related high-resource languages. In this paper, we evaluate the performance of these methods versus monolingual parsing of Tagalog, an Austronesian language which shares little typological similarity with any existing high-resource languages. We show that a monolingual model developed on minimal target language data consistently outperforms all cross-lingual and multilingual models when no closely-related sources exist for a low-resource language.- Anthology ID:
- 2020.udw-1.2
- Volume:
- Proceedings of the Fourth Workshop on Universal Dependencies (UDW 2020)
- Month:
- December
- Year:
- 2020
- Address:
- Barcelona, Spain (Online)
- Venue:
- UDW
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 8–15
- Language:
- URL:
- https://aclanthology.org/2020.udw-1.2
- DOI:
- Cite (ACL):
- Angelina Aquino and Franz de Leon. 2020. Parsing in the absence of related languages: Evaluating low-resource dependency parsers on Tagalog. In Proceedings of the Fourth Workshop on Universal Dependencies (UDW 2020), pages 8–15, Barcelona, Spain (Online). Association for Computational Linguistics.
- Cite (Informal):
- Parsing in the absence of related languages: Evaluating low-resource dependency parsers on Tagalog (Aquino & de Leon, UDW 2020)
- PDF:
- https://preview.aclanthology.org/auto-file-uploads/2020.udw-1.2.pdf