Towards Answering Climate Questionnaires from Unstructured Climate Reports

TL;DR

This paper introduces new datasets and models to extract structured information from unstructured climate reports, aiding climate change communication and policy-making.

Contribution

It presents large-scale climate questionnaire datasets, trains self-supervised models for text alignment, and establishes a benchmark for climate text classification.

Findings

Models generalize across organization types

Effective alignment of unstructured texts to questionnaires

Provides a new benchmark for climate NLP tasks

Abstract

The topic of Climate Change (CC) has received limited attention in NLP despite its urgency. Activists and policymakers need NLP tools to effectively process the vast and rapidly growing unstructured textual climate reports into structured form. To tackle this challenge we introduce two new large-scale climate questionnaire datasets and use their existing structure to train self-supervised models. We conduct experiments to show that these models can learn to generalize to climate disclosures of different organizations types than seen during training. We then use these models to help align texts from unstructured climate documents to the semi-structured questionnaires in a human pilot study. Finally, to support further NLP research in the climate domain we introduce a benchmark of existing climate text classification datasets to better evaluate and compare existing models.