Eyaa-Tom Dataset

Overview

Eyaa-Tom is a dataset designed to support Natural Language Processing (NLP) research for Togolese languages. It originates from the Eyaa-Tom dataset, meaning "People's Words" in Kabyè, which was collected through fieldwork by the Umbaji community of linguists, researchers, and annotators. The dataset focuses on speech and text data for various applications, including name identification and service queries in healthcare and finance, among other domains.

Dataset Details

• Total Languages Covered: 11 local languages + French with Togolese accent
• Included Data:
  • samples from Umbaji datasets for speech recognition in local languages (Moba, Nawdem, Ewe, etc.).
  • recordings per language featuring names in Togolese languages.

License & Access This dataset is owned by the Umbaji community. For full access, please contact:

📧 [email protected]

📧 [email protected]

Acknowledgments This dataset is the result of extensive field data collection efforts by the Umbaji community. We acknowledge the contributions of linguists, researchers, and annotators who made this resource possible.

Citation

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

@inproceedings{justin-etal-2025-yodiv3,
    title = "{Y}odi{V}3: {NLP} for {T}ogolese Languages with Eyaa-Tom Dataset and the Lom Metric",
    author = "Justin, Bakoubolo Essowe  and
      Xegbe, Kodjo Fran{\c{c}}ois  and
      Essuman, Catherine Nana Nyaah  and
      Samuel, Afola Kossi Mawou{\'e}na",
    editor = "Lignos, Constantine  and
      Abdulmumin, Idris  and
      Adelani, David",
    booktitle = "Proceedings of the Sixth Workshop on African Natural Language Processing (AfricaNLP 2025)",
    month = jul,
    year = "2025",
    address = "Vienna, Austria",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2025.africanlp-1.20/",
    doi = "10.18653/v1/2025.africanlp-1.20",
    pages = "143--149",
    ISBN = "979-8-89176-257-2",
    abstract = "Most of the 40+ languages spoken in Togo are severely under-represented in Natural Language Processing (NLP) resources. We present YodiV3, a comprehensive approach to developing NLP for ten Togolese languages (plus two major lingua francas) covering machine translation, speech recognition, text-to-speech, and language identification. We introduce Eyaa-Tom, a new multi-domain parallel corpus (religious, healthcare, financial, etc.) for these languages. We also propose the Lom metric, a scoring framework to quantify the AI-readiness of each language in terms of available resources. Our experiments demonstrate that leveraging large pretrained models (e.g.NLLB for translation, MMS for speech) with YodiV3 leads to significant improvements in low-resource translation and speech tasks. This work highlights the impact of integrating diverse data sources and pretrained models to bootstrap NLP for under-served languages, and outlines future steps for expanding coverage and capability."
}