Machine Learning for Climate Policy: Understanding Policy Progression in the European Green Deal

Abstract

This research applies machine learning techniques to analyse how climate policies progress from announcement through adoption within the European Green Deal framework. The team compiled data on 165 policies, testing various text representation methods including TF-IDF, BERT, and ClimateBERT alongside metadata features. Results show that ClimateBERT outperforms other approaches using text alone, while BERT achieves superior performance with the addition of metadata features. Explainable AI methods revealed the significance of policy language and institutional factors including political affiliation and geographic location in determining progression outcomes.

Peer Review

Community peer review and discussion of this publication.