Take part
Challenge Statement
The Metascience Novelty Indicators Challenge is a global initiative to develop and validate scalable indicators that can identify novelty in metascience. We are challenging academics, data scientists, researchers, and innovators worldwide to create tools that could reshape how we evaluate novelty and thereby support metascience progress.
The Metascience Novelty Indicators Challenge invites individuals or teams who can generate novelty scores for journal publications across a wide range of disciplines
Phases of the Challenge
-
Starting now
- Sign up for the interest list (below)
-
Provide expert judgment ratings (when invited)
- Meet our team at academic conferences, including:
Ongoing
Metascience community building and advancing the field of novelty in research. -
Phase 1: Expression of Interest (September 2025)
EOIs will open in September 2025
You can submit an application including:
- Personal details for you and other members of your team
- Description of your proposed approach to measuring novelty
- Evidence of your ability to deliver the indicator
- Commitment to the challenge timeline and requirements
-
Phase 2: Development Period (October 2025 – January 2026)
- Successful applicants will develop and refine their solutions between October and December
- Once indicators are more developed, participants will receive the testing data in January 2026 to test their indicators against expert human raters and then refine their tool
- You will have the opportunity to test your approach across multiple disciplines
-
Phase 3: Submission & Assessment (February – March 2026)
- Submit your indicator results alongside a methodology description
- Assessment will be based primarily on accuracy against our expert human ratings
- Consideration of cross-disciplinary applicability might be used.
-
Post-awarding
- Partnership opportunities to advance your new indicator
- Opportunity to contribute to a peer-reviewed publication reporting the various indicators used in the challenge
How to win
The winner of the £300,000 grand prize will be the challenger whose indicator best replicates human raters’ expert evaluation of research novelty.
Challengers will be given a list of tens of thousands of papers, spanning many fields and disciplines. They will use their preferred approach to generate a novelty score for each paper. Supplying this set of scores, along with a high-level description of the methods used, is all that is required for this challenge.
Entries will be tested against a uniquely large dataset produced on an unprecedented scale. The organising team are systematically collating independent human judgments of novelty by relevant experts across the same list of papers.
The precise winning criteria will be detailed for challengers in due course, but challengers will be assessed on the extent to which their scores match the human judgements.
Your models remain yours, and how/what you choose to publish about your methodology and code will be up to you. But, by entering the challenge, you will be agreeing to have your entry featured on a ‘leaderboard’ and in a major journal article about the challenge. You are free to publish separately about your entry.