The DOI registration agency Crossref has published a position paper on the use of persistent identifiers (PIDs) in scholarly infrastructure. It emphasizes the need for a holistic approach to PIDs, focusing on their metadata, services, interoperability, and long-term sustainability.
A persistent identifier is a stable designation whose value depends on the quality of the associated metadata and services. While a PID without quality metadata reveals the existence of an entity, it tells you nothing about its properties or context. Many types of identifiers are used in research, each fulfilling a different role at various stages of the research process. The identifier for digital objects, such as datasets or publications, is the DOI (Digital Object Identifier). Crossref, one of the 12 registration agencies of the DOI Foundation, operates a system that enables the assignment of resolvable identifiers with rich, open metadata—primarily for objects within the publishing process. However, it is important to remember that DOI metadata (and PIDs in general) are not homogeneous. For instance, records registered with one registration agency cannot be searched within the system of another. This creates an opportunity for national and international strategies, which—by setting clear expectations and requirements for registration agencies—can directly influence open research and foster its connectivity and sharing.
Key Positions:
- PIDs are a means to an end, not an end in themselves. The value of an identifier system lies in metadata enrichment, service scalability, and meeting the needs of the communities it serves.
- PIDs should complement, not compete with, one another. Identifiers can productively coexist as long as they serve their intended community and purpose.
- Not all DOIs are created equal. Different DOI registration agencies provide distinct services, metadata, and governance structures. Any policy or its implementation must therefore define specific requirements and expected outcomes for each identified entity, rather than focusing solely on the type of identifier.
- Interoperability is essential. Deploying identifier systems without assessing their interoperability undermines the very effort to connect research outputs. National strategies should therefore ensure that PID infrastructure is integrated into the broader, global research ecosystem.
- Sustainability and governance matter as much as technical standards. Any assessment of infrastructure organizations must take into account their community governance and demonstrated financial stability.
- Open infrastructure—not PID adoption alone—is the foundation of a connected, open research record. Strategies should measure success based on the openness, connectivity, and reach of the surrounding infrastructure and its services.
Recommendations for Policymakers and Implementers:
- Define goals. Focus on open research infrastructure and open metadata. Clarify what specific problem the PID is meant to solve and whether the surrounding infrastructure is built for that purpose.
- Be specific. Establish technical, operational, and governance requirements so that implementation—including the choice of a DOI registration agency—is evidence-based and delivers intended outcomes.
- Require more than just PID registration. The chosen PID system should enable the use of open, machine-readable metadata and its interconnection across the research ecosystem. It should also be governed to increase the reach of the record and ensure its long-term persistence.
- Evaluate identifier sustainability and governance.
- Build on existing open infrastructure. If considering a custom local system, carefully weigh the integration costs and the years required to achieve sufficiently broad community adoption.