Filling the Gap for Uzbek: Creating Translation Resources for Southern Uzbek
Not on the map. Why?
Abstract (arXiv)
Southern Uzbek (uzs) is a Turkic language variety spoken by around 5 million people in Afghanistan and differs significantly from Northern Uzbek (uzn) in phonology, lexicon, and orthography. Despite the large number of speakers, Southern Uzbek is underrepresented in natural language processing. We present new resources for Southern Uzbek machine translation, including a 997-sentence FLORES+ dev set, 39,994 parallel sentences from dictionary, literary, and web sources, and a fine-tuned NLLB-200 model (lutfiy). We also propose a post-processing method for restoring Arabic-script half-space characters, which improves handling of morphological boundaries. All datasets, models, and tools are released publicly to support future work on Southern Uzbek and other low-resource languages.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 8 of its 22 references that are in the map. First posted on arXiv 20 Aug 2025.
Builds on
8 of 22 references.
References · 8
- 2018SentencePiece: A simple and language independent subword tokenizer and detokenizer for Neural Text Processingcited by 1,989
- 2018A Call for Clarity in Reporting BLEU Scorescited by 1,799
- 2022No Language Left Behind: Scaling Human-Centered Machine Translationcited by 1,312
- 2018Adafactor: Adaptive Learning Rates with Sublinear Memory Costcited by 905
- 2020Language-agnostic BERT Sentence Embeddingcited by 627
- 2023MADLAD-400: A Multilingual And Document-Level Large Audited Datasetcited by 181
- 2022The first neural machine translation system for the Erzya languagecited by 7
- 2020Joint Persian Word Segmentation Correction and Zero-Width Non-Joiner Recognition Using BERTcited by 4
Built on by
Not cited yet.