UQA: Corpus for Urdu Question Answering

arXiv:2405.01458v1 Announce Type: new
Abstract: This paper introduces UQA, a novel dataset for question answering and text comprehension in Urdu, a low-resource language with over 70 million native speakers. UQA is generated by translating the Stanford Question Answering Dataset (SQuAD2.0), a large-scale English QA dataset, using a technique called EATS (Enclose to Anchor, Translate, Seek), which preserves the answer spans in the translated context paragraphs. The paper describes the process of selecting and evaluating the best translation model among two …

abstract anchor arxiv cs.ai cs.cl cs.ir cs.lg dataset english generated language low novel paper question question answering scale speakers stanford text translate type