- Location: Ireland, Northern Ireland
- Institution: Queen's University Belfast, University College Dublin
- Status: Active
- Type: Independent
- Theme: Evidence Discovery & Integration
- Timeframe: 2024 - 2029
Share on social
The volume of new research being published on climate, biodiversity, and water is growing faster than can realistically be analysed. Traditionally, pulling evidence together into useful insights for policymakers has been a slow, labour-intensive process.
This project aims to change that by training artificial intelligence (AI) models to read, interpret, and summarise scientific literature automatically – producing up-to-date evidence summaries in a fraction of the time, and giving researchers and decision-makers a clearer picture of what the evidence currently suggests.
The system will also indicate how confident we should be in its findings, similar to the approach used by the Intergovernmental Panel on Climate Change (IPCC), which aims to communicate how certain or uncertain its conclusions are. That transparency is essential when making decisions on complex environmental issues where the science is still evolving.
Project Goals
-
Develop a clear framework for using AI to automatically gather and summarise evidence from scientific literature.
-
Collect and organise large volumes of research related to climate change, biodiversity, and water quality, ensuring it is ready to be processed by AI.
-
Train and refine AI models specifically for the task of reading and synthesising scientific evidence.
-
Test the accuracy and reliability of the system by comparing its outputs against manually reviewed sources and IPCC reports, with a particular focus on areas where the science is uncertain or contested.
-
Connect the system with the tools and platforms that policymakers and researchers already use, making it as straightforward as possible to adopt.
-
Work with key stakeholders to gather feedback, refine the system, and ensure it is genuinely useful and accessible in practice.
Institutions
