CEntRE

CEntRE是一个针对企业关系提取的中文数据集，由中国科学院信息工程研究所创建。该数据集包含3018个段落，涵盖11018个关系三元组和10894个企业实体，涉及19种企业间关系。数据来源于公开的商业新闻数据，通过人工标注和智能数据处理构建。CEntRE的特点包括实体重叠、灵活的名称结构、跨句子和同义词推理以及广泛的应用场景覆盖。该数据集旨在解决企业关系提取的挑战，支持风险分析、评级研究和供应链安全等实际应用。CEntRE is a Chinese dataset for enterprise relation extraction, developed by the Institute of Information Engineering, Chinese Academy of Sciences. This dataset contains 3018 paragraphs, encompassing 11018 relation triples and 10894 enterprise entities, covering 19 types of inter-enterprise relationships. It is constructed from publicly available business news data through manual annotation and intelligent data processing. Key characteristics of CEntRE include entity overlap, flexible entity naming structures, cross-sentence and synonym reasoning, as well as broad coverage of application scenarios. This dataset is designed to address the challenges in enterprise relation extraction and supports practical applications such as risk analysis, rating research, and supply chain security.

搜集汇总

背景与挑战

背景概述

CEntRE是一个专门用于企业关系提取的中文数据集，由中国科学院信息工程研究所开发，包含3018个段落、11018个关系三元组和10894个企业实体，覆盖19种企业间关系。数据来源于公开商业新闻，通过人工标注和智能处理构建，具有实体重叠、灵活名称结构和跨句子推理等特点，旨在支持风险分析、供应链安全等实际应用。