Designing institutional services for persistent identifiers

Dr Maude Frances1, Dr Daniel Bangert1, Mr Harry Sidhunata1

1UNSW Sydney, Sydney, Australia


In an increasingly diverse landscape of scholarly outputs and infrastructure, persistent identifiers (PIDs) have become a key technology in tackling issues of accessibility, integration, and interoperability. For researchers, PIDs improve the visibility and citability of research, allow work to be more easily tracked and measured, and demonstrate a level of curation and governance over objects.

PID service providers include CrossRef, DataCite, ORCID, FundRef, ISNI, and Ringgold. Several of these organisations have worked to improve integration across international research infrastructure through collaborative projects such as the ORCID and DataCite Interoperability Network (ODIN) and Technical and Human Infrastructure for Open Research (THOR), and by organising meetings about the future of open identifiers. Persistent identifiers also feature in the work of research community organisations such as the Research Data Alliance and FORCE11.


At UNSW Library, implementation of PIDs has focused on DOIs and ORCID iDs, seeking to leverage the international efforts that have already seen widespread adoption of these content and contributor identifiers by the scholarly community. PIDs flow through a number of systems and tools, including UNSW’s publication management system (an installation of Symplectic Elements) and library repositories for publications, faculty-based collections, and research data (instances of Fedora repositories).

The UNSW DOI Service is a web application that enables UNSW researchers to request a DOI for grey literature that has been deposited in the institutional repository (UNSWorks) or a faculty-based repository [1] [2]. These research outputs include theses, conference papers, conference presentations, conference posters, reports, working papers, and creative works. DOIs are minted through the Australian National Data Service, a member of DataCite. DOIs minted in the service point to an existing persistent link (Handle), which resolves to the landing page for the record. The landing page displays metadata about the record and includes links to download the publication file.

The service allows the user to search for grey literature in the repository, review publication details, enter any missing mandatory metadata, and agree to conditions for requesting a DOI. Once a request is submitted, it is reviewed by an administrator (UNSW Library staff) who either mint a DOI or decline the request. When a DOI is minted, it is then added to the record metadata and if requested, the DOI is also added to the publication file.

An alternative workflow exists for ‘trusted partners’ who need to mint a large number of DOIs or are required to mint DOIs on a regular basis. For example, a faculty, school or centre administrator responsible for an ongoing series of reports. Trusted partners are given access to part of the web application that enables them to search for the relevant record and mint a DOI without external review (see Figure 1).


The UNSW DOI Service is integrated with a Citation Builder tool that uses descriptive metadata (DC or MODS) to display a citation in Harvard style. The tool can also be configured to draw on metadata in different schemas and display citations in alternative styles. The purpose of citation display is twofold: for the user to confirm details of the record prior to requesting a DOI, and to give the user a citation for use after a DOI has been minted.

Integrations with other identifiers include the use of Handles for resolving DOIs, and mapping ORCID identifiers for contributors to the DataCite metadata schema. Integration between identifiers for content (DataCite DOIs) and contributors (ORCID iDs) result in a number of benefits for discovery and reporting:


Figure 1: UNSW DOI Service workflow for trusted partners

  1. ORCID identifiers are drawn from UNSW’s publication management system, ensuring that the identifier has been verified by the researcher.
  • ORCID identifiers are displayed in the repository, making the author’s ORCID profile, including other works listed in their profile, more visible and discoverable.
  • Repository metadata, including ORCID identifiers, are harvested by external aggregators (e.g. Trove, BASE), further exposing the ORCID profile.
  • By including ORCID iDs within DataCite metadata, authorship is unambiguously identified in a machine-readable, structured manner.
  • Works with a DataCite DOI can be easily added to the author’s ORCID profile via manual claim or auto-update from DataCite to ORCID.


As emphasised by THOR, widespread use and trust in PIDs throughout the research lifecycle depends largely on the ‘human infrastructure’ offered by these services. At UNSW, key messages for the research community focus on how PIDs make research easier to find and cite. In addition, they help track research impact through citations and altmetrics. It is also anticipated that PIDs will improve submission and reporting workflows by becoming a tool for auto-populating information and claiming records from other systems. Examples of this include publishers and repositories that offer authentication via ORCID and reference managers that support DOI lookup.


The design and delivery of PID services at UNSW Library is guided by the characteristics of ‘trusted identifiers’: unique, persistent, descriptive, interoperable, and governed [3]. Identifiers assigned to UNSW research outputs are interoperable, based on institutional sources of truth, and contain human and machine-readable metadata. In the case of DOIs, persistent access to the resource is guaranteed by the library as the custodian of the identifiers and associated repository content. As identifiers become a part of each stage of the research lifecycle, the challenge for institutions will be to continue to effectively govern their assignment, follow standards set by community-driven efforts, and optimise their use for and by researchers.


  1. Bangert, D. & Frances, M. (2016, November). Cite my thesis? DOIs for grey literature. CAUL Research Repositories 2016, Sydney.
  2. Mitra, P., Sheng, M., Sidhunata, H., & Bangert, D. (2017). UNSW DOI Service[software]. Zenodo.
  3. ODIN Consortium, Fenner, M., Thorisson, G., Ruiz, S., & Brase, J. (2013). D4.1 Conceptual model of interoperability. figshare.


Dr Maude Frances leads the Library Repository Services unit in The University of New South Wales (UNSW Sydney) Library. She has led a number of research infrastructure projects in which the Library collaborated with researchers to develop workflows and services for curating research data and publications.