REFinD
收藏资源简介:
REFinD数据集是由庆熙大学的研究团队创建,专注于金融领域的实体关系提取。该数据集包含28,676个实例,涵盖22种关系,数据来源于美国证券交易委员会(SEC)的公开报告。创建过程中,研究团队利用了EDGAR系统和SEC的财务报表及注释数据集进行数据收集和预处理。REFinD数据集的应用领域主要集中在金融文本的自动化处理和分析,旨在通过提取实体间的关系,为金融决策提供支持。
The REFinD dataset was developed by a research team from Kyung Hee University, specializing in entity relation extraction within the financial domain. This dataset comprises 28,676 instances covering 22 distinct relation types, with its data sourced from public reports of the U.S. Securities and Exchange Commission (SEC). During the dataset construction process, the research team utilized the EDGAR system and SEC's financial statements and notes datasets for data collection and preprocessing. The primary application scenarios of the REFinD dataset focus on automated processing and analysis of financial texts, aiming to provide support for financial decision-making by extracting relations between entities.

- 1FinTree: Financial Dataset Pretrain Transformer Encoder for Relation Extraction庆熙大学 · 2023年



