£309k funding to explore how AI can identify breakthrough science

Dr Angelo Salatino, from the Open University's (OU) Knowledge Media Institute (KMi), has secured £308,702 (£186,028 to the OU) in funding from UK Research and Innovation (UKRI) for a project that aims to identify potentially transformative research earlier and more effectively.

Working in partnership with University of Pittsburgh, the University of Oxford and Elsevier, the project, Predicting Disruption and Novelty, will develop AI-powered indicators that can analyse scientific papers at the point of publication to predict their novelty and disruptive potential.

Current measures of scientific impact often rely on citation data, which can take years to accumulate. This makes it difficult for funders, institutions and policymakers to identify groundbreaking research in real time. The new project aims to address this challenge by using machine learning and Generative AI to detect signals of innovation directly from the content of research papers.

The team will also investigate the relationship between novelty and disruption in science. While novelty is often associated with unusual combinations of ideas, disruption reflects whether research changes the direction of future scientific activity. By examining both together, the project will explore why some new ideas strengthen existing knowledge while others create entirely new avenues of research.

The findings could help support more timely and scalable approaches to research assessment, improving how emerging breakthroughs are identified and understood across the scientific landscape.