Financial Reports Numerical Extraction (FINE)
收藏资源简介:
FINE数据集是由微软、北京大学和中国科学院软件研究所等机构共同创建的,专门用于从混合长文档中提取财务关键绩效指标(KPIs)。该数据集基于SEC的EDGAR系统,包含了大量财务报告数据,旨在解决大语言模型在处理混合文本和表格数据时的信息提取问题。数据集通过自动化信息提取框架(AIE)进行处理,能够有效支持财务分析等领域的应用。
The FINE Dataset was jointly created by institutions including Microsoft, Peking University, and the Institute of Software, Chinese Academy of Sciences, and is specifically designed for extracting financial key performance indicators (KPIs) from mixed long documents. Based on the SEC's EDGAR system, the dataset contains a large volume of financial report data, aiming to address the information extraction challenges faced by large language models (LLMs) when processing mixed text and tabular data. The dataset is processed via an automated information extraction framework (AIE), which can effectively support applications in fields such as financial analysis.

- 1Extract Information from Hybrid Long Documents Leveraging LLMs: A Framework and Dataset微软、北京大学、中国科学院软件研究所、蚂蚁集团 · 2024年



