{"ID":2836898,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2511.20182","arxiv_id":"2511.20182","title":"KyrgyzBERT: A Compact, Efficient Language Model for Kyrgyz NLP","abstract":"Kyrgyz remains a low-resource language with limited foundational NLP tools. To address this gap, we introduce KyrgyzBERT, the first publicly available monolingual BERT-based language model for Kyrgyz. The model has 35.9M parameters and uses a custom tokenizer designed for the language's morphological structure. To evaluate performance, we create kyrgyz-sst2, a sentiment analysis benchmark built by translating the Stanford Sentiment Treebank and manually annotating the full test set. KyrgyzBERT fine-tuned on this dataset achieves an F1-score of 0.8280, competitive with a fine-tuned mBERT model five times larger. All models, data, and code are released to support future research in Kyrgyz NLP.","short_abstract":"Kyrgyz remains a low-resource language with limited foundational NLP tools. To address this gap, we introduce KyrgyzBERT, the first publicly available monolingual BERT-based language model for Kyrgyz. The model has 35.9M parameters and uses a custom tokenizer designed for the language's morphological structure. To eval...","url_abs":"https://arxiv.org/abs/2511.20182","url_pdf":"https://arxiv.org/pdf/2511.20182v1","authors":"[\"Adilet Metinov\",\"Gulida M. Kudakeeva\",\"Gulnara D. Kabaeva\"]","published":"2025-11-25T11:05:53Z","proceeding":"cs.CL","tasks":"[\"cs.CL\"]","methods":"[\"Language Model\"]","has_code":false}