News Story

KMi’s contributions to Open Repositories 2025: Advancing Open Access and Research Innovation

Friday 4 Apr 2025

CORE will be contributing seven accepted submissions to the 20th International Conference on Open Repositories (OR2025), taking place in Chicago, Illinois, USA, from 15–18 June 2025. These presentations highlight ongoing efforts to enhance open access, improve research discoverability, and address key challenges in the open repositories community.

From managing machine access in the era of generative AI to improving research classification and repository interoperability, each submission provides valuable insights for repository managers, academic institutions, and the wider open access ecosystem. Highlights include:

Managing Access to Open Repositories in the Age of Generative AI: A panel discussion addressing the challenges repositories face as open access content is increasingly used for training large language models.
USRN Discovery Pilot: A presentation sharing results from a project that enhances research discoverability in US repositories by leveraging CORE as an indexing service.
SDG-Classify: An AI model that automatically classifies research papers according to UN Sustainable Development Goals (SDGs), integrated into the CORE Dashboard.
Describing Software in Record Metadata: A lightning talk exploring the challenges of representing open research software in metadata and introducing the SoFAIR project.
Open Access Repositories Tracking Project: A research project generating annual global statistics on open research papers to improve understanding of repository content.
Identifying and Extracting Data Access Statements: A presentation on a new machine learning module developed by CORE to automatically identify and extract Data Access Statements (DAS) from full-text research articles.
Interoperable Verification and Dissemination of Software Assets: A submission introducing the SoFAIR project, which leverages machine learning to extract software mentions from research papers to promote reproducible science.