Text Cleaning with Transformer Language Models for Hungarian
Acta Cybernetica · 02.06.2026
DOI: 10.14232/actacyb.316365 ↗Kurz gefasst
Fachartikel in der Zeitschrift Acta Cybernetica, online erschienen am 02.06.2026. Verfasst von Gábor Madarász und 4 weiteren Autor:innen. Die Arbeit nennt 25 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
In language technology, clean data is fundamental for training high-quality models, yet large corpora often contain substantial noise due to OCR errors, missing diacritics, and various user-generated inconsistencies. This paper presents a comprehensive text cleaning pipeline tailored for Hungarian, leveraging transformer-based language models optimized for three key tasks: OCR error correction, diacritic restoration, and filtering grammatically incorrect sentences. We introduce huT5, a Hungarian adaptation of the mT5 model, which reduces model parameters and resource demands while maintaining strong performance on Hungarian-specific text cleaning tasks. The huT5 models were fine-tuned on carefully constructed Hungarian corpora for each task and benchmarked against state-of-the-art methods, demonstrating competitive results, particularly in OCR error correction and diacritic restoration. Our pipeline offers an efficient, freely accessible solution to enhance data quality for Hungarian NLP applications, setting a new standard in resource-efficient, language-specific text cleaning.
Zitieren
APA
Madarász, G., Holl, A., Ligeti-Nagy, N., Yang, Z. G., & Váradi, T. (2026). Text Cleaning with Transformer Language Models for Hungarian. Acta Cybernetica. https://doi.org/10.14232/actacyb.316365
Vancouver
Madarász G, Holl A, Ligeti-Nagy N, Yang ZG, Váradi T. Text Cleaning with Transformer Language Models for Hungarian. Acta Cybernetica. 2026. doi:10.14232/actacyb.316365
BibTeX
@article{Madarsz2026,
author = {Madarász, Gábor and Holl, András and Ligeti-Nagy, Noémi and Yang, Zijian Győző and Váradi, Tamás},
title = {Text Cleaning with Transformer Language Models for Hungarian},
journal = {Acta Cybernetica},
year = {2026},
doi = {10.14232/actacyb.316365}
}
EUVIMED