clarin-pl/combo-nlp-xlm-roberta-base-turkish-boun-ud2.17

COMBO-NLP Model for Turkish

Model Description

This is a Turkish-language model based on combo-nlp, an open-source natural language preprocessing system. It performs:

• —sentence segmentation (via combo-seg)
• —tokenisation (via combo-seg)
• —part-of-speech tagging
• —morphological analysis
• —lemmatisation
• —dependency parsing

The Turkish model uses `FacebookAI/xlm-roberta-base` as its base encoder and is trained on UD_Turkish-BOUN (UD v2.17).

Evaluation

Evaluation was performed on the UDTurkish-BOUN test split using the standard [CoNLL 2018 eval script](https://universaldependencies.org/conll18/conll18ud_eval.py).

Two evaluation rows are reported:

• —Full-text (F1): raw text is segmented by combo-seg, then parsed and compared against gold — measures end-to-end pipeline performance including segmentation quality.
• —Aligned accuracy: accuracy on correctly segmented (aligned) tokens — measures parsing quality on tokens that were correctly identified by the segmenter.

Morphosyntactic Tagging

Dependency Parsing

Usage

Install the library from PyPI (assuming you have a virtual environment created):

pip install combo-nlp

The combo-seg segmenter (used to split and tokenise raw text) is installed automatically as a dependency of combo-nlp, so no extra install step is needed.

from combo import COMBO

# Load a pre-trained model with the corresponding combo-seg segmenter
nlp = COMBO("Turkish")

# Parse raw text (handles sentence splitting + tokenization)
result = nlp("Çevik kahverengi tilki tembel köpeğin üzerinden atlar.")

# Inspect results
for sentence in result:
    for token in sentence:
        print(f"{token.form:<15} {token.lemma:<15} {token.upos:<8} head={token.head}  {token.deprel}")

Refer to the combo-nlp documentation for installation and usage instructions:

• —https://pypi.org/project/combo-nlp/
• —https://pypi.org/project/combo-seg/

License

The training data license: cc-by-sa-4.0 is derived from the Universal Dependencies treebank. For the full license terms of each treebank, please refer to the corresponding LICENSE.txt file in the treebank repository:

• —UD_Turkish-BOUN LICENSE.txt

Citation

If you use this model, please cite:

Ulewicz, M., Jabłońska, M., Klimaszewski, M., Przybyła, P., Pszenny, Ł., Rybak, P., Wiącek, M., & Wróblewska, A. (2026). COMBO-NLP Models Trained on UD v2.17. Zenodo. https://doi.org/10.5281/zenodo.19650523

@software{combo_nlp_2026,
  author    = {Ulewicz, Michał and Jabłońska, Maja and Klimaszewski, Mateusz and Przybyła, Piotr and Pszenny, Łukasz and Rybak, Piotr and Wiącek, Martyna and Wróblewska, Alina},
  title     = {{COMBO-NLP} Models Trained on {UD} v2.17},
  year      = {2026},
  publisher = {Zenodo},
  doi       = {10.5281/zenodo.19650523},
  url       = {https://doi.org/10.5281/zenodo.19650523}
}

Treebank References

If you use the Turkish BOUN treebank data, please also cite:

@article{marcsan2022enhancements,
  title={Enhancements to the BOUN Treebank Reflecting the Agglutinative Nature of Turkish},
  author={Mar{\c{s}}an, B{\"u}{\c{s}}ra and Akkurt, Salih Furkan and {\c{S}}en, Muhammet and G{\"u}rb{\"u}z, Merve and G{\"u}ng{\"o}r, Onur and {\"O}zate{\c{s}}, {\c{S}}aziye Bet{\"u}l and {\"U}sk{\"u}darl{\i}, Suzan and {\"O}zg{\"u}r, Arzucan and G{\"u}ng{\"o}r, Tunga and {\"O}zt{\"u}rk, Balk{\i}z},
  journal={arXiv preprint arXiv:2207.11782},
  year={2022}
}

@article{TurkEtAl2022,
  title = {Resources for {{Turkish}} Dependency Parsing: Introducing the {{BOUN Treebank}} and the {{BoAT}} Annotation Tool},
  author = {T{\"u}rk, Utku and Atmaca, Furkan and {\"O}zate{\c s}, {\c S}aziye Bet{\"u}l and Berk, G{\"o}zde and Bedir, Seyyit Talha and K{\"o}ksal, Abdullatif and Ba{\c s}aran, Balk{\i}z {\"O}zt{\"u}rk and G{\"u}ng{\"o}r, Tunga and {\"O}zg{\"u}r, Arzucan},
  year = {2022},
  month = mar,
  journal = {Language Resources and Evaluation},
  volume = {56},
  number = {1},
  pages = {259--307},
  issn = {1574-0218},
  doi = {10.1007/s10579-021-09558-0}
}

Resources

• —combo-nlp: https://pypi.org/project/combo-nlp/
• —combo-seg: https://pypi.org/project/combo-seg/
• —UDTurkish-BOUN: [https://github.com/UniversalDependencies/UDTurkish-BOUN](https://github.com/UniversalDependencies/UD_Turkish-BOUN)