Building a Robust Research Infrastructure, One PID at a Time

Tom Demeranville's picture

Enabling a wide range of connections between ORCID iDs and other persistent identifiers (PIDs) is a key element of our strategic plan, vital to achieving our shared vision of a PID-enabled research infrastructure. But, to ensure that those connections are valuable to and trusted by the community, all identifiers in the ORCID Registry need to meet some basic requirements.

There are many types of identifiers and they offer different levels of utility.  At a basic level, a PID is exactly what you’d imagine -- a reference to a person, place, or thing, which can be used to uniquely identify them, in perpetuity. PIDs may be internal (i.e., for use within a single organization); proprietary (for use within a single system); or open (fully interoperable in any system). You probably won’t be surprised that, at ORCID, we like open PIDs the best, since they are the easiest to work with for everyone. However, we also welcome the use of proprietary PIDs, as long as they resolve to enough information to help determine uniqueness (see below), and can be shared under a CC0 license in our data files. Internal PIDs can also be added to ORCID records and shared in our data files, but we only allow them to be categorized as ”Non-standard ID from work data source” -- our way of saying internal accession number.

There are several other desirable characteristics which make some PIDs more useful for making trusted connections -- or assertions -- than others:

  • Resolvable PIDs: These are either URLs (links), or can be transformed into URLs, which  resolve directly to a document or a human-readable landing page using well-known rules. Generally ORCID expects this category of PIDs to also provide machine-readable metadata, but that is outside the strictest definitions.  Example: Requests For Comments (RFC) are assigned a PID. The ORCID Registry can use this PID to generate links to the webpage containing the RFC.  The identifier “rfc6750” becomes https://tools.ietf.org/html/rfc6750.
  • FAIR PIDs: These PIDs are not just resolvable, but can also be used to discover open, interoperable, well-defined metadata containing provenance information in a predictable manner.  They are openly governed for the benefit of the community. Example: DOIs are stored either as URLs “https://doi.org/10.1/123”, or simply “10.1/123”.  We present these to the user as links in the Registry and you can also follow those links to discover metadata describing the linked item.  DOIs are governed by the International DOI Foundation and the attached metadata is available under a CC0 license, meaning that it is open to everyone. The metadata contains information about the publisher, the publication, other authors, funding, and affiliation(s), all of which help establish the provenance of the item.  Other FAIR PIDs include arXiv identifiers, PubMed and PubMed Central identifiers and most ISBN identifiers.

As a benefit of membership, organizations can ask ORCID to support additional PID types in the ORCID Registry.  For example, a member could request that we add support for the PIDs they use for identifying samples or datasets in their geology database.  This enables links between the samples and the people who collected them. And, because all new PID types that we add must be at least resolvable, and preferably FAIR, those links are unambiguous, persistent over time, and actionable - benefitting the researcher, the member organization, and the wider community

Adding Your PID!

As a community organization, we want to ensure that ORCID supports the PIDs used by our members. We maintain a complete list of existing identifiers supported in the Registry, and invite ORCID members to use this form to request additional PIDs. We aim to respond to your request within 48 hours; please allow two weeks for your PID to be added.

Thank you for helping us build a robust research infrastructure, one PID at a time!