zzunyang/KLQD

KLQD（韩国法律问答数据集）是一个针对法律AI应用的大规模数据集，包含了来自NAVER知识iN、韩国法律援助公司（KLAC）和易于查找实用法律（EFPL）三个主要平台的法律问答数据。该数据集共有16,419对问答，涵盖了包括民法、刑法、家庭法、劳动法、行政法在内的广泛法律主题。数据集通过结构化的预处理流程确保高质量，包括文本清洗、去重和法律领域过滤。此外，通过结合基于GPT-4的自动标注和人工验证，法律问题被分类到九个主要法律领域。KLQD为法律AI模型训练提供了一个多样化的、专家验证的资源，有助于AI应用在法律理解和推理方面的改进。

KLQD (Korean Legal QA Dataset) is a large-scale dataset for legal AI applications, which consists of legal Q&A data collected and processed from three major platforms: NAVER Knowledge iN, Korea Legal Aid Corporation (KLAC), and Easy to Find Practical Law (EFPL). The dataset includes 16,419 QA pairs covering a wide range of legal topics such as civil, criminal, family, labor, and administrative law. It ensures high-quality data through a structured preprocessing pipeline including text cleaning, deduplication, and legal domain filtering. Furthermore, legal questions are categorized into nine primary legal fields through a combination of GPT-4-based automatic annotation and human verification. KLQD provides a diverse, expert-verified resource for legal AI model training, facilitating improved reasoning and legal understanding in AI applications.