Paper Type

ERF

Abstract

Environmental, Social, and Governance (ESG) data are increasingly vital for strategic decision-making, yet remains fragmented and inconsistently reported across sources. Existing ESG analytics solutions typically address structuring, retrieval, or summarization in isolation, limiting their ability to capture complex, multi-entity relationships. We propose a unified ESG data processing framework that integrates knowledge graph construction, community detection, and Graph-based Retrieval-Augmented Generation (GraphRAG) to deliver structured, context-rich insights. Our system models interconnected ESG entities, applies network analysis to uncover hidden patterns, and generates traceable, decision-ready responses to complex queries. We evaluate an early implementation using a dataset of 8,350 ESG reports from 309 companies, demonstrating improvements in coverage, contextual relevance, and multi-perspective analysis. To address potential biases in ESG disclosures, we outline plans for cross-source validation and sentiment analysis. Our contributions establish a scalable and transparent foundation for ESG analytics, offering enhanced traceability, diversity, and decision support for analysts, regulators, and stakeholders.

Paper Number

1850

Author Connect URL

https://authorconnect.aisnet.org/conferences/AMCIS2025/papers/1850

Comments

SIGDSA

Author Connect Link

Share

COinS
 
Aug 15th, 12:00 AM

Enhancing Environmental, Social, and Governance Insights with Structured Knowledge Graphs and Advanced Query Analysis

Environmental, Social, and Governance (ESG) data are increasingly vital for strategic decision-making, yet remains fragmented and inconsistently reported across sources. Existing ESG analytics solutions typically address structuring, retrieval, or summarization in isolation, limiting their ability to capture complex, multi-entity relationships. We propose a unified ESG data processing framework that integrates knowledge graph construction, community detection, and Graph-based Retrieval-Augmented Generation (GraphRAG) to deliver structured, context-rich insights. Our system models interconnected ESG entities, applies network analysis to uncover hidden patterns, and generates traceable, decision-ready responses to complex queries. We evaluate an early implementation using a dataset of 8,350 ESG reports from 309 companies, demonstrating improvements in coverage, contextual relevance, and multi-perspective analysis. To address potential biases in ESG disclosures, we outline plans for cross-source validation and sentiment analysis. Our contributions establish a scalable and transparent foundation for ESG analytics, offering enhanced traceability, diversity, and decision support for analysts, regulators, and stakeholders.

When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.