FinTruthQA
收藏资源简介:
FinTruthQA是由浙江大学等机构创建的金融信息披露质量评估基准数据集,包含6000条来自上海和深圳证券交易所投资者互动平台的真实金融问答数据。该数据集通过人工标注,依据问题识别、问题相关性、答案可读性和答案相关性四个维度进行评估。创建过程中,研究者解决了自然语言解释、上下文敏感性和主观性评估等技术挑战。FinTruthQA旨在通过自动化评估提升金融报告的透明度和质量,适用于审计、监管和金融分析等领域,以增强投资者信心和市场效率。
FinTruthQA is a benchmark dataset for financial information disclosure quality assessment, developed by institutions such as Zhejiang University. It contains 6,000 real financial question-and-answer (Q&A) pairs sourced from the investor interaction platforms of the Shanghai Stock Exchange and Shenzhen Stock Exchange. Manually annotated, this dataset is evaluated across four dimensions: question identification, question relevance, answer readability, and answer relevance. During its development, researchers addressed technical challenges including natural language interpretation, context sensitivity, and subjective assessment. FinTruthQA aims to enhance the transparency and quality of financial reports through automated assessment, and is applicable to fields such as auditing, regulation, and financial analysis, with the goal of boosting investor confidence and market efficiency.




