17 October 2022

RAM AI at the Sustainable Finance Hack 2022

The Sustainable Finance Hack is a 24h project development event organised by Open Geneva and the Geneva School of Economics and Management – UNIGE and aims to unite tech and sustainability enthusiasts to brainstorm and prototype solutions for sustainable finance.

Detecting controversies related to ESG (Environmental, Social and Governance) is essential when evaluating and improving sustainability within investment portfolios. Training a machine learning model is an efficient method to detect such controversies in an automatic and data-driven way. This year, RAM AI hosted the challenge “Build a Bad Buzz Factory for ESG Controversy Detection!” and was motivated by the observation that ESG-specific data may sometimes be difficult or expensive to collect, which may impact research advances and applications of data-driven detection methods.

Committed to solving this issue, we were joined by a team of data scientists, developers, and master students, to explore the idea of generating data using large language models to train and enhance machine learning-based detection models. The preliminary evaluation showed promising results, encouraging the team to continue the collaboration after the hackathon. We believe the outcome of this project will benefit the community and facilitate the development of data-driven methods for ESG-related research.