Machine Translation Between High-resource Languages in a Language Documentation Setting. Kann, K., Ebrahimi, A., Stenzel, K., & Palmer, A. In Proceedings of the first workshop on NLP applications to field linguistics, pages 26–33, Gyeongju, Republic of Korea, October, 2022. International Conference on Computational Linguistics. ![Machine Translation Between High-resource Languages in a Language Documentation Setting [link]](https://bibbase.org/img/filetypes/link.svg "link")
Paper abstract bibtex Language documentation encompasses translation, typically into the dominant high-resource language in the region where the target language is spoken. To make data accessible to a broader audience, additional translation into other high-resource languages might be needed. Working within a project documenting Kotiria, we explore the extent to which state-of-the-art machine translation (MT) systems can support this second translation – in our case from Portuguese to English. This translation task is challenging for multiple reasons: (1) the data is out-of-domain with respect to the MT system's training data, (2) much of the data is conversational, (3) existing translations include non-standard and uncommon expressions, often reflecting properties of the documented language, and (4) the data includes borrowings from other regional languages. Despite these challenges, existing MT systems perform at a usable level, though there is still room for improvement. We then conduct a qualitative analysis and suggest ways to improve MT between high-resource languages in a language documentation setting.
@inproceedings{kann_machine_2022,
address = {Gyeongju, Republic of Korea},
title = {Machine {Translation} {Between} {High}-resource {Languages} in a {Language} {Documentation} {Setting}},
url = {https://aclanthology.org/2022.fieldmatters-1.3},
abstract = {Language documentation encompasses translation, typically into the dominant high-resource language in the region where the target language is spoken. To make data accessible to a broader audience, additional translation into other high-resource languages might be needed. Working within a project documenting Kotiria, we explore the extent to which state-of-the-art machine translation (MT) systems can support this second translation – in our case from Portuguese to English. This translation task is challenging for multiple reasons: (1) the data is out-of-domain with respect to the MT system's training data, (2) much of the data is conversational, (3) existing translations include non-standard and uncommon expressions, often reflecting properties of the documented language, and (4) the data includes borrowings from other regional languages. Despite these challenges, existing MT systems perform at a usable level, though there is still room for improvement. We then conduct a qualitative analysis and suggest ways to improve MT between high-resource languages in a language documentation setting.},
urldate = {2023-01-09},
booktitle = {Proceedings of the first workshop on {NLP} applications to field linguistics},
publisher = {International Conference on Computational Linguistics},
author = {Kann, Katharina and Ebrahimi, Abteen and Stenzel, Kristine and Palmer, Alexis},
month = oct,
year = {2022},
pages = {26--33},
}
Downloads: 0
{"_id":"zAzYBsCvPdnuZNqCG","bibbaseid":"kann-ebrahimi-stenzel-palmer-machinetranslationbetweenhighresourcelanguagesinalanguagedocumentationsetting-2022","author_short":["Kann, K.","Ebrahimi, A.","Stenzel, K.","Palmer, A."],"bibdata":{"bibtype":"inproceedings","type":"inproceedings","address":"Gyeongju, Republic of Korea","title":"Machine Translation Between High-resource Languages in a Language Documentation Setting","url":"https://aclanthology.org/2022.fieldmatters-1.3","abstract":"Language documentation encompasses translation, typically into the dominant high-resource language in the region where the target language is spoken. To make data accessible to a broader audience, additional translation into other high-resource languages might be needed. Working within a project documenting Kotiria, we explore the extent to which state-of-the-art machine translation (MT) systems can support this second translation – in our case from Portuguese to English. This translation task is challenging for multiple reasons: (1) the data is out-of-domain with respect to the MT system's training data, (2) much of the data is conversational, (3) existing translations include non-standard and uncommon expressions, often reflecting properties of the documented language, and (4) the data includes borrowings from other regional languages. Despite these challenges, existing MT systems perform at a usable level, though there is still room for improvement. We then conduct a qualitative analysis and suggest ways to improve MT between high-resource languages in a language documentation setting.","urldate":"2023-01-09","booktitle":"Proceedings of the first workshop on NLP applications to field linguistics","publisher":"International Conference on Computational Linguistics","author":[{"propositions":[],"lastnames":["Kann"],"firstnames":["Katharina"],"suffixes":[]},{"propositions":[],"lastnames":["Ebrahimi"],"firstnames":["Abteen"],"suffixes":[]},{"propositions":[],"lastnames":["Stenzel"],"firstnames":["Kristine"],"suffixes":[]},{"propositions":[],"lastnames":["Palmer"],"firstnames":["Alexis"],"suffixes":[]}],"month":"October","year":"2022","pages":"26–33","bibtex":"@inproceedings{kann_machine_2022,\n\taddress = {Gyeongju, Republic of Korea},\n\ttitle = {Machine {Translation} {Between} {High}-resource {Languages} in a {Language} {Documentation} {Setting}},\n\turl = {https://aclanthology.org/2022.fieldmatters-1.3},\n\tabstract = {Language documentation encompasses translation, typically into the dominant high-resource language in the region where the target language is spoken. To make data accessible to a broader audience, additional translation into other high-resource languages might be needed. Working within a project documenting Kotiria, we explore the extent to which state-of-the-art machine translation (MT) systems can support this second translation – in our case from Portuguese to English. This translation task is challenging for multiple reasons: (1) the data is out-of-domain with respect to the MT system's training data, (2) much of the data is conversational, (3) existing translations include non-standard and uncommon expressions, often reflecting properties of the documented language, and (4) the data includes borrowings from other regional languages. Despite these challenges, existing MT systems perform at a usable level, though there is still room for improvement. We then conduct a qualitative analysis and suggest ways to improve MT between high-resource languages in a language documentation setting.},\n\turldate = {2023-01-09},\n\tbooktitle = {Proceedings of the first workshop on {NLP} applications to field linguistics},\n\tpublisher = {International Conference on Computational Linguistics},\n\tauthor = {Kann, Katharina and Ebrahimi, Abteen and Stenzel, Kristine and Palmer, Alexis},\n\tmonth = oct,\n\tyear = {2022},\n\tpages = {26--33},\n}\n\n\n\n","author_short":["Kann, K.","Ebrahimi, A.","Stenzel, K.","Palmer, A."],"key":"kann_machine_2022","id":"kann_machine_2022","bibbaseid":"kann-ebrahimi-stenzel-palmer-machinetranslationbetweenhighresourcelanguagesinalanguagedocumentationsetting-2022","role":"author","urls":{"Paper":"https://aclanthology.org/2022.fieldmatters-1.3"},"metadata":{"authorlinks":{}}},"bibtype":"inproceedings","biburl":"https://bibbase.org/zotero/abhishek-p","dataSources":["h7kKWXpJh2iaX92T5"],"keywords":[],"search_terms":["machine","translation","between","high","resource","languages","language","documentation","setting","kann","ebrahimi","stenzel","palmer"],"title":"Machine Translation Between High-resource Languages in a Language Documentation Setting","year":2022}