Connected Open Source Software
Researchers require simple – and ideally automated – pathways to generate high-quality, FAIR-compliant metadata for research software. In practice, however, a lack of time, resources or specialist expertise often hinders researchers from keeping metadata up to date on a consistent basis. Existing tools and schemas frequently fall short, leaving metadata inconsistent or incomplete and offering limited options for automation.
ConnOSS addresses these challenges by establishing an infrastructure that offers a central hub for research software while harmonising and enriching metadata from a variety of sources and presenting it in a machine-readable format. Built on GitHub and GitLab, the platform aggregates metadata consistently and publishes it via open web infrastructures.
Using machine learning, metadata is automatically extracted from structured and unstructured sources such as repository APIs, README files and citation data and enhanced as necessary. Interoperability and reusability are guaranteed by a custom metadata schema based on Schema.org, which integrates established standards including CodeMeta, Bioschemas and maSMP.
The purpose of the project is to improve the visibility, FAIRness and reproducibility of research software while supporting good scientific practice. With minimal additional effort, researchers can document their software in a FAIR-compliant manner, enabling it to be reused over the long term – both by humans and by machines.
ZB MED’s role in the project
- Project coordination
- Helping to define metadata standards
- Supporting metadata extraction and harmonisation
- Implementing dissemination and outreach activities
Duration
- 1 September 2025 - 31 August 2028
Funding body
- German Research Foundation (DFG) - Scientific Library Services and Information Systems (DFG-LIS): funding programme "Research Software Infrastructure", grant number 561044496
Partners
-
Carl von Ossietzky University of Oldenburg
-
GESIS - Leibniz Institute for the Social Sciences
