Persistent identifiers are designed to provide stable references to entities and to keep knowledge connected across systems, yet in practice there may be justified restrictions on which data can be publicly exposed. This creates a tension between persistence, transparency, and access control.
In the MetaBelgica project, four Belgian Federal Scientific Institutes collaboratively manage and connect cultural heritage reference data across institutional boundaries in an internal Wikibase instance. This talk presents two complementary perspectives. First, it shows how an off-the-shelf PID infrastructure can be realised by combining Wikibase with its stable entity identifiers and built-in mechanisms for handling change, with Archival Resource Key (ARK) identifiers as resolvable external identifiers.
Second, it reflects on how this approach is adapted in a setting with restricted data, where only selected entities and properties are exposed while internal metadata remains hidden. This highlights practical design choices around identifier resolution and access control.
How do you keep collaboratively curated knowledge connected when reference data is evolving and not fully open? In many real-world systems, some entities or properties must remain internal for good reasons, such as workflow management or legal constraints. This raises practical questions for PID design: what should an identifier resolve to, and how do you ensure persistence when not all data can be publicly exposed?
In this session you will discover how a combination of Wikibase and Archival Resource Key identifiers can provide a practical, off-the-shelf PID infrastructure for collaboratively maintained reference data. Wikibase assigns stable identifiers (Q numbers) to entities and handles changes such as edits and merges through built-in mechanisms like revision histories and redirects. ARK identifiers can be layered on top to provide resolvable, persistent identifiers that can be shared externally. This includes the actual PID, but possibly also revisions of the entity. Together, these components already offer much of what is needed for PID infrastructure that helps keep identifiers consistent and knowledge connected across institutional systems, without requiring extensive custom development.
Building on this foundation, the session then explores how this approach is applied in the MetaBelgica project, where four Belgian Federal Scientific Institutes work together on a shared knowledge base to connect and maintain cultural heritage reference data in an internal Wikibase. In this context, not all data can be made public. Visibility is controlled at both entity and property level, meaning that only selected information is exposed through the public interface. As a result, the implementation deliberately does not expose full revision histories through ARK resolution, even though Wikibase provides them internally.
You will see how this affects practical design choices, including what identifiers resolve to, how restricted entities are handled, and how to balance persistence with access control. The session highlights what works out of the box, what requires adaptation, and where limitations arise when applying PID principles in a partially restricted environment.
After this session you will:
see how Wikibase Q numbers provide stable internal identifiers for collaboratively edited entities
learn how ARK identifiers can be layered on top to provide external persistent identifiers
explore how combining Wikibase and ARKs can form an off-the-shelf PID infrastructure, even when some data must remain restricted
This session is relevant if you work with evolving reference data and need to implement persistent identifiers in a context where openness and access control must be carefully balanced.
Sven Lieber works as data manager at The Royal Library of Belgium (KBR), where he supports the improvement of data quality, mainly by using computational methods and Linked Data. Sven is currently involved in the research infrastructure project MetaBelgica and the research project BELTRANS. For both projects, Sven helps to make data available according to the FAIR principles. He studied Computer Science at the University of Freiburg in Germany and pursued a PhD related to Knowledge Graphs at the University of Ghent in Belgium, where he was also involved in several interdisciplinary research projects at the IDLab research group