Skip to content

saturnMars/derivingStructuredInsightsFromSustainabilityReportsViaLargeLanguageModels

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

17 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Glitter or Gold? Deriving Structured Insights from Sustainability Reports via Large Language Models

by Marco Bronzini, Carlo Nicolini, Bruno Lepri, Andrea Passerini and Jacopo Staiano

The minimal implementation of the paper "Glitter or Gold? Deriving Structured Insights from Sustainability Reports via Large Language Models". Published in the EPJ Data Science journal.

Abstract

Over the last decade, several regulatory bodies have started requiring the disclosure of non-financial information from publicly listed companies, in light of the investors’ increasing attention to Environmental, Social, and Governance (ESG) issues. Publicly released information on sustainability practices is often disclosed in diverse, unstructured, and multi-modal documentation. This poses a challenge in efficiently gathering and aligning the data into a unified framework to derive insights related to Corporate Social Responsibility (CSR). Thus, using Information Extraction (IE) methods becomes an intuitive choice for delivering insightful and actionable data to stakeholders. In this study, we employ Large Language Models (LLMs), In-Context Learning, and the Retrieval-Augmented Generation (RAG) paradigm to extract structured insights related to ESG aspects from companies’ sustainability reports. We then leverage graph-based representations to conduct statistical analyses concerning the extracted insights. These analyses revealed that ESG criteria cover a wide range of topics, exceeding 500, often beyond those considered in existing categorizations, and are addressed by companies through a variety of initiatives. Moreover, disclosure similarities emerged among companies from the same region or sector, validating ongoing hypotheses in the ESG literature. Lastly, by incorporating additional company attributes into our analyses, we investigated which factors impact the most on companies’ ESG ratings, showing that ESG disclosure affects the obtained ratings more than other financial or company data.

KEYWORDS: ESG Dimensions, Non-financial Disclosures, Large Language Models, In-context Learning, Knowledge Graphs, Bipartite Graph Analyses, Interpretability

Supplementary Material document

Repository structure

Data

This folder includes the sustainability reports, and our inputs, in their original format (PDFs) and after text extraction (textual files).

Code

This folder includes the minimal code to extract structured data from textual documents and conduct the analyses reported in the paper.

Output

Here the triples extracted using our methodology represent the ESG-related actions disclosed by the selected companies.

Availability of ESG data

Unfortunately, due to copyright and licence issues, I cannot disclose any data related to companies' ESG scores, extracted from the Refinitiv platform. Accessing such a platform requires a costly subscription and its terms & conditions forbid the re-distribution. Conversely, I understand the importance of ESG-related data and apologise for not helping your research.

Some universities or research institutes (especially those focus on economics) might have a licence for research only purposes, providing researchers with such valuable data sources. On the other hand, independent researchers could have serious difficulties to gain access to it. I would however suggest exploring further data sources such as those listed here.

About

Minimal code for extracting structured Insights from Sustainability Reports via Large Language Models

Topics

Resources

Stars

Watchers

Forks