Mr Angelo Salatino

Scientific progress and innovation depend on a complex interplay between researchers, funding, teams, time and ideas. Funding agencies, including UKRI, balance a choice between investing in people, as the cornerstone of a robust UK research workforce, with investing in research ideas underpinning potentially commercial innovations. For example, whereas previous research has worked to identify relevant markers for high-performing research individuals (stars) [citeme], it is more difficult to identify indicators of market success early in the innovation pipeline. This increases the risk associated with framing an entire funding system around early investment in potential research ideas, when the vast majority of these will fail to materialise [citeme]. Furthermore, early investment in ideas as sole focus, risks underinvesting in a workforce that is necessary to underpin the development of novel ideas in the long term. This project labels this paradox as a choice between funding research “stars” (individual research careers), versus research “unicorns” (ideas or concepts). What is needed is a funding mechanism that invests in people and the sustainability of a research workforce, but also provides an environment for novel, and potentially commercial ideas, can thrive. This project leverages Natural Language Processing (NLP), and Knowledge Graphs extracting concepts in the titles and abstracts. And it matches them with bibliometric data, monitoring of research success on individuals who were both successful and unsuccessful to attract UKRI research funding. It will result in the development of an AI-driven database capable of track-ing the evolution of ideas and careers, over time. This database, known as the TRACK-database (Trajectories of Research Applications and Career Knowledge), will be the world’s first longitudinal funding database linked to pseudonymised individuals and their academic output capable of mapping UKRI-funded and unfunded grant applications to individual researchers, research concepts, and innovation over time. Creation of this database represents a step change in research policy evaluation by introducing a scalable, AI-powered framework that will enable large-scale, AI-driven analyses. These include analysing how careers and ideas develop over time, as well as causally linking these longitudinal outcomes with different funding types (fellowship or teams) at the early stage of a research career (<5 years first publication). Further investigations include whether certain disciplines benefit more from specific funding models and assess the UK's position in the global research landscape by comparing its research trends with international competitors. Engagement with UKRI, as a major beneficiary of this project’s outcomes, will ensure that research insights from the TRACK-database are utilised towards the development and implementation of actionable insights used to facilitate a UK funding environment that contributes to both a sustainable UK workforce, as well as fosters the development of novel ideas and concepts. Such policies will be used to develop a more resilient and productive research ecosystem, thereby enhancing UK’s international competitiveness.

The researcher will join a team dedicated to developing a custom analysis of AI bibliometric trends. Specifically, Angelo will consult with the AI Index team on designing a classification scheme for AI publications to be featured in the 2025 AI Index report. Once the scheme is established, there is also interest in publishing the analysis as a standalone research paper. While the AI Index team will conduct the analysis, Angelo’s role will be consultative, providing strategic guidance on project design.