Open, Findable, Connected: Persistent Identifiers for Research

This article was written with the support of Miguel Rey Mazon. He is a research assistant and member of the RDM team at TU Graz Library and Archives.

What are Persistent Identifiers?

Persistent Identifiers (PIDs) are universal, unique and permanent strings of characters that can be used to identify physical and digital objects, such as publications, data sets and research instruments, as well as concepts, such as people, places, time periods and events, on the web. These character strings appear in the form of a URI (Uniform Resource Identifier) and refer to the metadata of a described entity (in some cases, a digital object), enabling meaningful reuse. Metadata is 'data about data' that characterises an object (such as a digital publication) on a descriptive, administrative and technical level.

Unlike conventional URLs, which only point to a location on the web where the corresponding content can be found, PIDs identify the content itself. This can be either a pure metadata record or a digital object accompanied by metadata. Even if the content — be it a research dataset or a digital publication — is moved to another location, it can still be found thanks to its persistent identification. In the event of a change of location, the entity's metadata is updated, but the identifier remains the same. This dynamic functionality prevents links from 'breaking' and thus contributes to the long-term findability of data on the internet.

PIDs are becoming increasingly important due to developments in recent years towards an open research landscape. They actively support the FAIR principles (Findable, Accessible, Interoperable, Reusable), which specify several guidelines for improving the reusability of (research) data.

The PID Landscape - A Short Summary

While persistent findability and reusability sound promising at first, they must also be considered in the right context. No system lasts forever, and even PIDs can break, albeit less frequently than URLs. To ensure data consistency, a great deal of work is required to maintain and manage the individual systems. The concept of persistently referencing physical objects was already in practice before the 'digital turn' through standards such as the ISBN number or various identifiers for biological samples. In this context, however, the focus is on implementing the concept using digital technologies.

The PID landscape comprises many different standards. These include intradisciplinary standards, such as CETAF (Consortium of European Taxonomic Facilities) in biology, as well as interdisciplinary standards (e.g., DOI, URN) that are applied across various disciplines and entities. Additionally, there are identifiers that refer to single entities such as individuals or organisations (e.g., ORCID, ROR). The choice of a PID scheme for objects in research processes depends on the size and resources of the institution using it, as well as on national conditions.

In the research environment, the following standards have become established at a European level: ORCID (Open Researcher and Contributor ID) for individuals; DOI (Digital Object Identifier) or URN (Uniform Resource Name) for publications and datasets; and ROR (Research Organization Registry) for organisations. The final version of the RAiD (Research Activity Identifier) identifier for research projects was published in 2025 and is still being developed towards system compatibility. In some cases, the GND (Gemeinsame Normdatei) is also used to reference individuals, institutions and publications. In recent years, DOIs have become dominant in scientific publications, due to the publishing landscape.

Each PID is linked to a specific schema which defines the identifier's structure. For example, DOIs are part of the Handle system, while the ORCID schema has evolved from the ISNI (International Standard Name Identifier) schema.

Example DOI: The administration and management of PIDs is a very complex process. While so-called 'authorities' (e.g. the DONA Foundation) are responsible for the stability and persistence of the systems, the assignment and resolution of PIDs is often handled by 'service providers' such as DataCite or CrossRef.

As a DataCite member, the TU Wien Library has offered the DOI Service Austria since 2020. The Library acts as a local registration agency, supporting Austrian universities, research institutions and other non-profit organisations in registering and using DOIs. A consortium for ORCID was also formed at TU Wien in 2019. The consortium's goal is to establish the standard throughout Austria. TU Graz is a member of the ORCID consortium and uses the national services to register DOIs. The DOIs in the institutional repository are therefore assigned via a DataCite interface.

How can PIDs benefit Research?

To achieve compliance with the open science strategy, it is necessary to permanently identify resources linked to research processes for sustainable and transparent research management. However, PIDs' true value lies in their ability to make the research process and all its entities comprehensible, and to track developments. PIDs, such as DOIs, can be assigned to various entities throughout the research cycle — including publications, research data, software and measuring instruments — enabling them to be linked to each other, as well as to researchers and organisations.

Avoiding Ambiguities

Using PIDs and metadata enables objects to be uniquely identified, preventing confusion and incorrect assignments. Identical names among researchers are no longer an issue, as the metadata provides information on their scientific background and organisational affiliation. Name changes due to marriage can also be adequately represented in this way.

Findable and Citeable

PIDs make data more discoverable and accessible in the research process, thereby increasing its visibility. In addition, they enable individual entities to be cited by linking to basic metadata. This metadata can be reused by other platforms, such as BASE and OpenAIRE, and facilitates the import of publications into reference management programmes.

For example, our institutional research information system, PURE, has an integrated connection to ORCID, enabling researchers at Graz University of Technology to import their stored publications into the FIS with no extra effort.

FAIR Research Data Management

The implementation of the FAIR principles is being promoted at a European level through projects such as the European Open Science Cloud (EOSC). This is changing the requirements for project proposals and data management plans (DMPs). In many cases, researchers must now provide a researcher ID (e.g. an ORCID) and assign PIDs (e.g. DOIs) to their research outputs.

Another step in this direction is to link multiple PIDs in the spirit of Linked Open Data, making relationships between individual entities visible. This allows the entire research workflow to be clearly represented, from projects (RAiD) and researchers (ORCID) to institutions (ROR) and publications and research data (DOI).

Critical Voices

Although the technical requirements for implementing PIDs are in place and they are used by many institutions, some disadvantages (still) prevent their widespread use.

In some cases, such as with ORCID, PIDs can be registered free of charge. However, the situation is somewhat different for PIDs for digital objects, such as publications. The field is dominated by proprietary providers who are responsible for 'minting' the identifiers, as well as system maintenance and management. For example, to assign DOIs to digital objects, you must be a member of either DataCite or CrossRef, paying an annual membership fee as well as a flat fee for each assigned DOI. Such costs are often beyond the means of smaller institutions. To support Austrian institutions, TU Wien is a member of the DataCite association. As a local registration agency, it enables manual and automated registration at the national level, and provides technical support.

There are also non-proprietary formats for digital objects (e.g., ARK) that initially involve lower costs. However, considerable additional effort is required in terms of management, hosting, administration and distribution. This do-it-yourself character, which demands a certain level of technical expertise, has likely hindered the adoption of this standard. Nevertheless, it is being adopted mainly by cultural heritage institutions in France and Brazil.

Conclusion and Outlook

PIDs support good scientific practice and facilitate the management of scientific output, associated entities and work steps. The integration of scientific information into a variety of systems is supported by the functionalities of linked (open) data. Furthermore, they play a key role in the FAIR development of the research landscape.

While the PID landscape and its key players are already in place, as are recommendations from individual institutions on the use of PIDs in research, widespread application to all entities in the research process has not yet been achieved. Smaller institutions often struggle with the associated costs, and individual PIDs for entities within a research workflow are still being developed (e.g. RAiD, identifier for software).

However, PIDs are already standard practice in research repositories for identifying publications and individuals. Projects such as PID4NFDI in Germany are currently working on further developing national and international infrastructures. Some metadata standards are also being updated to meet the new requirements for the appropriate integration of PIDs (e.g. the Europeana Data Model). Further best practice examples for handling PIDs are likely to emerge in the coming years.

Further Information

• RDM Team – FAIR Data
• GO FAIR Austria
• National PID Services
• PIDs for Research Data (forschungsdaten.info, ger)
• PID Monitor (Germany)