Low-resource text classification using cross-lingual models for bullying detection in the ukrainian language

Скорочена інформація про документ
dc.contributor.authorOliinyk, V.
dc.contributor.authorMatviichuk, І.
dc.date.accessioned2023-05-16T08:43:38Z
dc.date.available2023-05-16T08:43:38Z
dc.date.issued2023
dc.description.abstractThis paper aims on building bullying detection model for Ukrainian language. Considering absence of labeled datasets for bullying detection and classification in Ukrainian, small Ukrainian dataset (4k samples) was gathered and used for testing models in this research. Taking into account very small number of Ukrainian datasets in general this dataset is publicly available for testing and benchmarking other text classification models. Modern approaches to text class classification in low-resource languages are studied in the paper. We apply zero-shot technique and evaluate performance of modern multilingual, cross-lingual state-of-the-art models and embeddings for text classification in Ukrainian language, including mBERT, XLM-R, LASER and MUSE. Experimental results shows that zero-shot approaches for classification task allow to achieve F1 score of 67-69% for multilingual models trained on English dataset only, having 88-91% test accuracy on English data. We also show that machine translation of English data can be used for estimating model performance in other languages, i.e. only 0-2% difference in test accuracy compared to natural data was received for best models XLM-R and LASER. Zero-shot approach for binary detection task showed even better results 81% compared to 91,59% on original English data. We then enhance the best XLM-R model by training it on our natural Ukrainian dataset and confirm benefits of augmenting low-resource language dataset with machine transla tions from resource-rich English data. Finally, the model for bullying detection in the Ukrainian language is built achieving F1 score of 91,59% with only 12k samples dataset in different languages.uk
dc.format.pagerangePp. 87-100uk
dc.identifier.citationOliinyk, V. Low-resource text classification using cross-lingual models for bullying detection in the ukrainian language / V. Oliinyk, І. Matviichuk // Адаптивні системи автоматичного управління : міжвідомчий науково-технічний збірник. – 2023. – № 1 (42). – С. 87-100. – Бібліогр.: 24 назви.uk
dc.identifier.doihttps://doi.org/10.20535/1560-8956.42.2023.279093
dc.identifier.urihttps://ela.kpi.ua/handle/123456789/55725
dc.language.isoenuk
dc.publisherКПІ ім. Ігоря Сікорськогоuk
dc.publisher.placeКиївuk
dc.relation.ispartofАдаптивні системи автоматичного управління : міжвідомчий науково-технічний збірник, 2023, № 1 (42)uk
dc.subjectmultilingual modelsuk
dc.subjectzero-shot classificationuk
dc.subjectbullying detectionuk
dc.subjectXLM-RoBERTauk
dc.subjectmBERTuk
dc.subjectLASERuk
dc.subjectMUSEuk
dc.subject.udc004.852uk
dc.titleLow-resource text classification using cross-lingual models for bullying detection in the ukrainian languageuk
dc.typeArticleuk

Файли

Контейнер файлів
Зараз показуємо 1 - 1 з 1
Ескіз
Назва:
279093-643342-1-10-20230512.pdf
Розмір:
806.57 KB
Формат:
Adobe Portable Document Format
Опис:
Завантажити
Ліцензійна угода
Зараз показуємо 1 - 1 з 1
Ескіз недоступний
Назва:
license.txt
Розмір:
9.1 KB
Формат:
Item-specific license agreed upon to submission
Опис:
Завантажити

Зібрання

Адаптивні системи автоматичного управління : міжвідомчий науково-технічний збірник. – 2023. – № 1 (42)