sfrontull/la-usc_valbadia_loresmt24

Dataset Card: Ladin (Val Badia) - Monolingual (La Usc di Ladins)

Overview

Source Paper: "Rule-Based, Neural and LLM Back-Translation: Comparative Insights from a Variant of Ladin"

Description: This dataset contains monolingual sentences in Ladin (Val Badia) from the newspaper "La Usc di Ladins" (https://www.lausc.it), which has been archived since 2008. The newspaper offers texts in five variants of Ladin, corresponding to the five Ladin valleys. We extracted 1,937,608 sentences from its archive and classified them according to variants using a specially trained classifier. Using this variant classifier, we filtered out only the sentences for the Val Badia variant. In addition, sentences with outdated spelling were removed, resulting in a clean corpus of 274,665 sentences. This refined dataset was used as back-translation data in the cited work. In the present data set, 3,697 sentences with telephone numbers and e-mail addresses were removed in addition to the original filtering.

Dataset Structure

Columns:

• text: Contains Ladin sentences.
• source: Indicates the source of each sentence (https://www.lausc.it/).

Format

• File Type: Parquet
• Encoding: UTF-8

Citation

If you use this dataset, please cite the following paper:

@inproceedings{frontull-moser-2024-rule,
    title = "Rule-Based, Neural and {LLM} Back-Translation: Comparative Insights from a Variant of {L}adin",
    author = "Frontull, Samuel  and
      Moser, Georg",
    editor = "Ojha, Atul Kr.  and
      Liu, Chao-hong  and
      Vylomova, Ekaterina  and
      Pirinen, Flammie  and
      Abbott, Jade  and
      Washington, Jonathan  and
      Oco, Nathaniel  and
      Malykh, Valentin  and
      Logacheva, Varvara  and
      Zhao, Xiaobing",
    booktitle = "Proceedings of the The Seventh Workshop on Technologies for Machine Translation of Low-Resource Languages (LoResMT 2024)",
    month = aug,
    year = "2024",
    address = "Bangkok, Thailand",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2024.loresmt-1.13",
    pages = "128--138",
    abstract = "This paper explores the impact of different back-translation approaches on machine translation for Ladin, specifically the Val Badia variant. Given the limited amount of parallel data available for this language (only 18k Ladin-Italian sentence pairs), we investigate the performance of a multilingual neural machine translation model fine-tuned for Ladin-Italian. In addition to the available authentic data, we synthesise further translations by using three different models: a fine-tuned neural model, a rule-based system developed specifically for this language pair, and a large language model. Our experiments show that all approaches achieve comparable translation quality in this low-resource scenario, yet round-trip translations highlight differences in model performance.",
}