Datasets:
language:
- uk
- en
license: cc-by-sa-4.0
task_categories:
- translation
pretty_name: SimIdioms — Aligned Ukrainian–English Idiom Corpus
tags:
- idioms
- phraseology
- ukrainian
- low-resource
- multilingual
- figurative-language
size_categories:
- 1K<n<10K
configs:
- config_name: default
data_files:
- split: train
path: sim_idioms.jsonl
SimIdioms
The first aligned Ukrainian–English idiom corpus — 2,262 clusters with idiom strings, translations, contextual example sentences, and figurative meanings on both language sides.
Companion to the paper "SimIdioms: A Corpus and Benchmark for Ukrainian Idiom Translation" (UNLP 2026). Code and evaluation framework: github.com/petrunivyaryna/sim-idioms.
Quick start
from datasets import load_dataset
ds = load_dataset("KSE-RESEARCH-Group/sim-idioms", split="train")
print(ds[0])
Schema
Each row is one idiom cluster:
{
"id_": "SIM-0",
"uk": {
"idiom": [{"text": "...", "field": "idiom" | "translation"}, ...],
"example": [{"text": "..."}, ...],
"figurative_meaning": [{"text": "..."}, ...],
"figurative_meaning_translated": [{"text": "..."}, ...]
},
"en": { ... same fields ... }
}
Total: 2,262 clusters. 3,595 EN / 2,751 UK idiom strings, 2,885 EN / 2,737 UK example sentences, 1,723 EN / 1,193 UK figurative meanings (original + translated).
License
The dataset is tagged CC BY-SA 4.0, inherited from the English-side
content (MIDAS). However, the Ukrainian-side content has additional
restrictions that override CC BY-SA for that portion — see LICENSE for the
full text.
- 🇬🇧 English-side content (derived from MIDAS) and the cluster-graph structure are released under CC BY-SA 4.0 — attribution + ShareAlike apply.
- 🇺🇦 Ukrainian-side content (idioms, figurative meanings, usage examples) derived from the Ukrainian-English and English-Ukrainian Phraseological Dictionary (Condor Publishing House) and the Dictionary of Phraseological Units of the Ukrainian Language (Bilonozhenko et al., 2003) is included with explicit publisher permission for academic research only. It may not be redistributed, mirrored, or used commercially without separate permission from Condor Publishing House.
By using this dataset you agree to use the Ukrainian-side content only for non-commercial academic research.
Citation
If you use this dataset, please cite our paper:
@inproceedings{petruniv-2026-simidioms,
title = {SimIdioms: A Corpus and Benchmark for Ukrainian Idiom Translation},
author = {Petruniv, Yaryna and Makogon, Iuliia and Kyslyi, Roman},
booktitle = {Proceedings of the Fourth Ukrainian Natural Language Processing Workshop (UNLP)},
year = {2026}
}
Please also cite the underlying sources
Because SimIdioms is built from existing resources, please also cite them when using the dataset, in line with their licenses (CC BY-SA 4.0 for the ShareAlike-licensed sources, and the Condor / Bilonozhenko publishers for the Ukrainian dictionary content).
@book{horot-etal-2024-ukrainsko,
title = {Ukrainsko-anhliiskyi ta anhlo-ukrainskyi frazeolohichnyi slovnyk
[Ukrainian-English and English-Ukrainian Phraseological Dictionary]},
author = {Horot, Ye. I. and Hromyk, Yu. V. and Malimon, L. K.
and Pavlenko, L. P. and Pavliuk, A. B. and Rohach, O. O.},
year = {2024},
publisher = {Condor Publishing House},
address = {Kyiv, Ukraine},
isbn = {978-617-7939-03-9}
}
@book{bilonozhenko2003slovnyk,
title = {Slovnyk frazeolohizmiv ukrainskoi movy
[Dictionary of Phraseological Units of the Ukrainian Language]},
author = {Bilonozhenko, V. M. and Hnatiuk, I. S. and Diatchuk, V. V.
and Nerovnia, N. M. and Fedorenko, T. O.},
publisher = {Naukova Dumka},
address = {Kyiv},
year = {2003}
}
@inproceedings{kim-etal-2025-memorization,
title = {Memorization or Reasoning? Exploring the Idiom Understanding of {LLM}s},
author = {Kim, Jisu and Shin, Youngwoo and Hwang, Uiji and Choi, Jihun
and Xuan, Richeng and Kim, Taeuk},
booktitle = {Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing},
year = {2025},
publisher = {Association for Computational Linguistics},
url = {https://aclanthology.org/2025.emnlp-main.1099/}
}
@inproceedings{chaplynskyi2023,
title = {Introducing {U}ber{T}ext 2.0: A Corpus of Modern {U}krainian at Scale},
author = {Chaplynskyi, Dmytro},
booktitle = {Proceedings of the Second Ukrainian Natural Language Processing Workshop (UNLP)},
year = {2023},
publisher = {Association for Computational Linguistics},
url = {https://aclanthology.org/2023.unlp-1.1/}
}