hassan7272/urdu-finance-qa

这是一个关于巴基斯坦和伊斯兰金融的乌尔都语金融问答数据集，包含1,510个精心策划的问题-答案对，涵盖了巴基斯坦用户的金融场景，如伊斯兰金融、银行、预算和数字支付。数据集的特点是包含了三种语言风格：纯乌尔都语、罗马乌尔都语和混合乌尔都语+英语，适合生产级的NLP系统。数据集特别关注伊斯兰金融合规性，如Zakat计算和资格、避免Riba（利息）等。每个条目包含问题、答案、类别、难度等信息。数据集的生成过程结合了大型语言模型和人工努力，确保了数据的真实性和一致性。

This is a Urdu financial QA dataset focused on Pakistan and Islamic finance, containing 1,510 carefully curated question-answer pairs covering financial scenarios relevant to Pakistani users, such as Islamic finance, banking, budgeting, and digital payments. The dataset features three linguistic styles: pure Urdu, Roman Urdu, and mixed Urdu+English, making it highly suitable for production-grade NLP systems. It particularly emphasizes Islamic financial compliance, including Zakat calculation and eligibility, and avoidance of Riba (interest). Each entry includes questions, answers, categories, difficulty levels, and more. The dataset was generated using large language models combined with manual effort to ensure realism and consistency.