With a growing global focus on corporate sustainability, responsible investment, and impact investing, the importance of sustainable information disclosure and sustainability ratings for corporate resource allocation and strategy is on the rise. This paper addresses the challenge faced by regulatory bodies, rating agencies, and companies in effectively measuring and enhancing ESG plan effectiveness. The purpose of this study is to employ machine learning to deconstruct the internal rules, embedded in the ESG rating results by rating agencies. Using predictive models with high accuracy, combined with Explainable Artificial Intelligence (XAI), the study aims to infer the relationships among varied metrics underlying ESG rating methodologies. The research further introduces a visual ESG dashboard for benchmark analysis, providing companies and their industry peers with in-depth insights into various sustainability topics and detailed rating outcomes. This innovative approach is designed to aid in assessing sustainability performance and shaping future strategies.

Author: Chia-Ming Sun

Published in: International Conference on Information Society (i-Society-2024)

• Date of Conference: 26-28 August, 2024
• DOI: 10.20533/iSociety.2024.0003
• ISBN: 978-1-913572-72-3
• Conference Location: Churchill College, Cambridge, UK