LR-Sum: Summarization for Less-Resourced Languages
Authors
Findings of the Association for Computational Linguistics: ACL 2023
pp. 6829-6844
July 1, 2023
Abstract
Introduces LR-Sum, an openly licensed multilingual summarization dataset containing human-written summaries for 40 languages. Kurdish is represented as separate Kurmanji and Sorani varieties. The paper establishes extractive and abstractive baselines, including mT5-based summarization, for less-resourced languages.
Keywords
Multilingual Summarization
Kurdish
Kurmanji
Sorani
mT5
Low-Resource NLP
Related Datasets
LR-Sum
Date not verified
40 languages; Kurdish represented as separate Kurmanji and Sorani subsets
Text, Article-summary pairs
A multilingual summarization dataset derived from Voice of America newswire with human-written summaries for 40 languages. It includes separate Kurmanji and Sorani Kurdish subsets and is released …
Citation
Palen-Michel, C., & Lignos, C. (2023). LR-Sum: Summarization for Less-Resourced Languages. Findings of the Association for Computational Linguistics: ACL 2023, 6829-6844.