Loading datasets...
License:
CC-BY-4.0
Steward:
CommunityDataset ID:
cmroyatmb00wko007rt1g1hr2
Release Date: 7/17/2026
Format: TSV
Size: 74.14 KB
Share
This dataset contains human-authored parallel sentence pairs for Ladino (Judeo-Spanish) aligned with English and Turkish. It is organised into seven configurations totalling 1,918 rows, yielding 3,333 Ladino-English and 845 Ladino-Turkish pairs. The original sentences were created and translated by the Sephardic Center of Istanbul (SKAD), with recollection and packaging done by Col·lectivaT. Since the data is genuine Ladino produced by the community (not synthetically generated), it is suitable for both training and evaluation. Created as part of the "Judeo-Spanish: Connecting the two ends of the Mediterranean" project.
Licensing
Creative Commons Attribution 4.0 International (CC-BY-4.0)
Restrictions/Special Constraints
Attribution must be given to the original creators (Sephardic Center of Istanbul and Col·lectivaT).
Forbidden Usage
Use without proper attribution to the original creators.
Ethical Review
The sentences were authored and translated by the Sephardic Center of Istanbul (SKAD), a cultural institution working on Sephardic/Judeo-Spanish heritage. The data consists of published/curated text and does not contain personal or sensitive information about identifiable individuals.
Intended Use
Training and evaluating machine translation models for Ladino, an endangered language, paired with English and Turkish. As the data is human-authored, it is suitable for evaluation as well as training.
This dataset was developed as part of the project "Judeo-Spanish: Connecting the two ends of the Mediterranean" carried out by Col·lectivaT and Sephardic Center of Istanbul. The original Ladino sentences and their English/Turkish translations were created by the Sephardic Center of Istanbul (SKAD); recollection and packaging were done by Col·lectivaT.
The data is split into seven configurations, each provided as a separate TSV file with an id column and the language-text columns for that configuration (eng, lad, tur):
guler.tur-eng-lad — 52 trilingual pairs (English-Ladino-Turkish)
onlineshop.eng-lad — 550 English-Ladino pairs
onlineshop.tur-lad — 17 Ladino-Turkish pairs
pandemi.lad-tur — 9 Ladino-Turkish pairs
tatoebalike.eng-lad — 997 English-Ladino pairs
tercumeler.eng-lad — 167 English-Ladino pairs
tercumeler.tur-lad — 126 Ladino-Turkish pairs
Unlike synthetically generated resources, this is genuine Ladino as produced by the community, making it suitable for both training and evaluation.
For other datasets resulting from this project visit the Ladino Data Hub. A live translation application is available at translate.sefarad.com.tr.
HuggingFace dataset repo: collectivat/skad_ladino_parallel
If you use this data, please cite:
Preparing an Endangered Language for the Digital Age: The Case of Judeo-Spanish
@inproceedings{oktem-etal-2022-preparing,
title = {Preparing an Endangered Language for the Digital Age: The Case of Judeo-Spanish},
author = {Öktem, Alp and Zevallos, Rodolfo and Moslem, Yasmin and Öztürk, Güneş and Şarhon, Karen},
booktitle = {Proceedings of the Workshop on Resources and Technologies for Indigenous, Endangered and Lesser-resourced Languages in Eurasia (EURALI)},
year = {2022},
url = {https://aclanthology.org/2022.eurali-1.18/}
}
Curation of this dataset was done as part of project "Judeo-Spanish: Connecting the two ends of the Mediterranean" carried out by Col·lectivaT and Sephardic Center of Istanbul within the framework of the "Grant Scheme for Common Cultural Heritage: Preservation and Dialogue between Turkey and the EU–II (CCH-II)" implemented by the Ministry of Culture and Tourism of the Republic of Turkey with the financial support of the European Union. The content of this repository is the sole responsibility of Col·lectivaT and does not necessarily reflect the views of the European Union.
Please check README.md for more information.