ESG-FTSE

ESG-FTSE数据集是由帝国理工学院、谷歌DeepMind和社会经济研究机构联合创建的，包含3,913篇英文新闻文章，涵盖2018年末至2021年夏季的时间段。该数据集专注于FTSE 100指数中市值排名前十的公司，旨在通过新闻文章分析公司的环境、社会和治理（ESG）相关性。数据集的创建过程包括数据收集、标注和评估，采用三层标注体系：二元分类、ESG分类和目标公司。ESG-FTSE数据集的应用领域主要集中在可持续投资研究，旨在通过自然语言处理技术提高ESG评分的准确性，从而推动负责任的投资决策。

The ESG-FTSE dataset was jointly created by Imperial College London, Google DeepMind, and socio-economic research institutions. It comprises 3,913 English news articles spanning the period from late 2018 to summer 2021. This dataset focuses on the top 10 companies by market capitalization in the FTSE 100 Index, aiming to analyze the environmental, social, and governance (ESG) relevance of enterprises through news articles. The dataset's creation process includes data collection, annotation and evaluation, adopting a three-tier annotation framework: binary classification, ESG classification, and target company annotation. The ESG-FTSE dataset is primarily applied in sustainable investment research, with the objective of enhancing the accuracy of ESG ratings using natural language processing (NLP) technologies, thereby facilitating responsible investment decision-making.

搜集汇总

背景与挑战

背景概述

ESG-FTSE数据集是一个专注于环境、社会和治理（ESG）分析的英文新闻数据集，包含2018年至2021年间约3,900篇关于FTSE 100指数中市值前十公司的文章。它采用三层标注体系，旨在通过自然语言处理技术提升ESG评分的准确性，主要应用于可持续投资研究领域。