All news at TU Wien

TU Wien goes TU Graz for the 2026 InvenioRDM partners workshop

Max Moser from the Center for RDM led sessions on the improvement of the REST API documentation and data curation workflows.

Group photo with 28 people.

© Dietmar Herbst

Each year, the InvenioRDM community gathers in one place to discuss what’s been happening and the future direction of the project. This year’s gathering was held at the campus of TU Graz (Graz University of Technology), for the full week of February 23rd to the 27th.

Since the institutional data repository of TU Wien is famously based on InvenioRDM and TU Wien is also a partner, participation was, of course, a no-brainer.

CERN & friends: usability improvements

Organizationally, Monday and Tuesday were dedicated to presentations given by CERN and other community members about the latest developments and some future work. This included the new AIRDEC project, over the course of which the deposit form will be entirely overhauled, to fix existing usability issues and build in more assistance for users uploading data.

Starting from Wednesday, the rest of the workshop was again organised with open space technology. This means that anybody could pitch ideas for topics that they deem relevant. Afterwards, every participant would vote on the topics that should be discussed in dedicated parallel sessions.

TU Wien: REST API and quality assurance

From TU Wien, Max Moser has pitched and led two sessions: one about improvement of the REST API documentation, and another about quality assurance and data curation workflows.

The earlier session was inspired by feedback from one of our researchers about the existing API documentation, Julia Jaklin, who used the REST API to upload over a hundred datasets to the E-LAUTE community. The session’s participants identified several pain points with the current centralised and manually maintained documentation, and provided ideas on how to alleviate them.

Overall, it was decided that the most promising way forward is to have each InvenioRDM instance’s documentation be built into the instance itself instead of centrally hosted. Having each instance automatically generate its OpenAPI specs should improve completeness and accuracy. Combined with customizable sets of explainers and examples, most pain points should be addressed.

Better review processes

Our second session explored the needs, workflows, and possible solutions for quality assurance that many institutions have. Several InvenioRDM partners are using the Invenio-Curations extension that was developed collaboratively by TU Graz and TU Wien. Others are using a simpler communities-based approach, with different workflows and use cases.

It was generally agreed that centralised teams of reviewers can easily become strained when the system becomes more popular. Also, generic reviewers may lack the domain-specific knowledge necessary to give the best feedback and suggestions possible to uploaders.

As such, TU Wien explained our push for faculty-based data stewards that addresses those issues. We also discussed the various paths we can take to enable suitable workflows in InvenioRDM generally and TU Wien Research Data in particular.

As it turns out, suitable mechanisms already exist in (the latest unreleased development version of) InvenioRDM that can be used to implement our requirements. Making these mechanisms available to our system and extending them to fit our use cases perfectly should be quite manageable and straightforward.

Many more fruitful discussions

Of course, there were many more fruitful discussions and sessions than the ones described above. A non-exhaustive listing includes improvements regarding the containerised setup, input and display of geospatial metadata, and better test suites for making accidental regression bugs less likely.

Contact

TU Wien
Center for Research Data Management
research.data@tuwien.ac.at