Insights into Academic Output of Universities of Applied Sciences using Linked Open Data

Universities of Applied Sciences (UAS) increasingly contribute to applied research and innovation, yet their scholarly outputs and collaborations are challenging to evaluate due to the fact that project, funding, and publication data are dispersed throughout various sources and often lack consistent identifiers. This article describes a Linked Open Data pipeline that combines (i) OpenAlex publication metadata and (ii) NWO/SIA project data that is exposed as RDF to create a project-centric research knowledge graph. The pipeline retrieves open-access PDFs, and uses a large language model to extract grant identifiers from full texts in order to link projects to publications. Starting from 52,640 UAS-affiliated publications in OpenAlex, 557 could be linked to a project record (1.1% of the total), indicating considerable coverage limitations even under"open" conditions. The pipeline demonstrates that linking projects to outputs via funding acknowledgements depends on the availability of open full texts and on consistent reporting of grant identifiers in publications. Although open scholarly graphs make it easier to find and analyse research on a large scale, they may underrepresent UAS research when outputs occur outside conventional publication infrastructures or when funding metadata is inconsistent. This article explains why grant-identifier-based linking by itself might not be enough for institutional research analysis and provides a reproducible method for linking publications and outputs. It also motivates broader strategies that combine open scholarly graphs with additional output sources and richer research metadata to support more representative institutional insight.