This repo serves as the code-base which underlines all the work pertianing to Climate Policy Transformer: Utilizing NLP to track the coherence of Climate Policy Documents in the Context of the Paris Agreement paper presented at ACL2024 Climate Change and AI sub-group. The repo contians two major dir:
- Dataset: This repo has all the codebase used to process the Climate Watch Sectoral Data For more details on why, please visit the paper link provided above.
- Classifiers: Climate Watch team has set-up a taxonomy based on which the climate policy documents are analyzed. The classiifers are trianed on these taxonomy to identify if the paragraph belongs to these catogories or not.
- RAG (coming soon): We have used RAG where the Retirever step is replaced by Classifiers to perform the relevant information extraction from policy documents.