The Climate Narrative Review: Decoding Environmental News: An NLP Pipeline for Climate Media Analysis
This capstone project, developed in collaboration with QuotaClimat, addresses a critical blind spot in environmental advocacy: the disconnect between the scale of environmental issues and how they are actually covered in the news. While media framing shapes public understanding and political will around climate change, NGOs and civil society organisations rarely have the tools to systematically track and analyse that framing at scale. Manual monitoring is time-consuming and inconsistent, leaving advocates without a reliable picture of where coverage falls short, which frames dominate, and where public narratives diverge from environmental reality.
To address this, the team built an NLP-based pipeline that automatically extracts structured insights from environmental news reports, identifying the frames present in coverage and surfacing discrepancies between environmental impact and media attention. Rather than requiring manual analysis, the tool processes news at scale and presents its findings in a form that supports informed, targeted action by NGOs and advocacy organisations. By turning unstructured media content into structured, interpretable data, the pipeline gives civil society a practical instrument for holding media accountable and strengthening the evidence base behind environmental communication strategies.