Guides to Choosing Persistent Identifiers - your feedback wanted!

The FREYA Project team is pleased to announce the publication of our initial guides on choosing persistent identifiers for different entities and we would like your feedback.

Published on Zenodo, these short guides were created to help with choosing persistent identifiers for various types of entities.
While not intended to be fully comprehensive, they are designed to provide a starting point for anyone thinking about using persistent identifiers in their systems. We created guides for the following entities:

  • Publications
  • Datasets
  • People
  • Organizations
  • Software

A table summarising all of the guides is also available.

We would value feedback and comments on these first versions throughout June 2020. You can provide feedback in the comments below or via email through Revised versions will be developed in July 2020.


Dear Frances,

Thank you for sharing these guides :slightly_smiling_face:

Some feedback about the People identifiers guide:

Page 1. The second branch of the tree “To identify individuals within my systems” only mentions ISNI, but actually ORCID can be (and is) used for this purpose. Integrating ORCID iDs and APIs in local systems can help solve the name ambiguity issue and supports interoperability.

Page 2. ORCID Format. The canonical form of the ORCID iD: ORCID. More info under

I hope this helps!


1 Like

These are great! Just shared on Twitter :slight_smile:

Hey Gabi, Re: point 1 - absolutely - the euroCROS/OCLC survey of research information systems found that 73% of the organisations that they surveyed used ORCIDs, and the majority of those will be using the API. See for details.

1 Like

Thanks for your comment, Josh!

Dear all, thank you for this guide.
I would like to suggest a few necessary add-ons to Publication identifiers.
ISSN has been used since its inception to identify proceedings, journal titles and more recently blogs.
One can easily find these various types of continuing resources on the ISSN portal easily (
ISSN is interoperable and cross-check data with CrossRef and Wikidata but also with a variety of information providers.
The ISSN portal also provides free metadata which is sufficient to identify titles. For subscribers, the ISSN portal provides reliable and up-to-date full metadata.
Access and searching are regular features of the ISSN portal for both humans and machines, thanks to our API. The ISSN Portal also provides information in linked data formats.
Moreover, all ISSN are embedded in URIs as described in ISO 3297 standard. ISSN URIs have the following syntax: This feature allows external databases to directly link to the page containing metadata.
Moreover, on the second page of the guide regarding Publication identifiers, these identifiers should be introduced in alphabetical order.

Thank you for amending this guide accordingly.
All the best,

Gaelle Bequet, ISSN IC Director


These guides look really helpful and we will definitely be using them at Imperial, thank you!

One comment, on the publication guide there is only reference to Crossref DOIs. Is it possible to include reference to DataCite or other DOIs that organisations (such as us) are using for publications?

Kind regards
Robyn Price (Imperial College London)

1 Like

Dear Frances, Hope this finds you well. Any reaction to my comments? Would be much appreciated. All the best, Gaelle

Hi Gaelle
Thanks to you and others for your useful comments. We will be reviewing all the comments received together in a few weeks and hope to publish new versions before the end of the summer.
If we have any queries about any of your suggestions we will be in touch.
Thanks again!

Hi Frances, many thanks for the update. Do not hesitate to contact me. All my best. Gaelle

Thanks for compiling these, very useful overviews! I like that they are written from the perspective of what functionality is conveyed by identifiers to the objects.

From the perspective of the actors choosing identifiers, it might be useful to include, in addition, characteristics as to costs of using the identifiers (if any), and perhaps also governance and business model of the organizations providing the identifiers (as is included for organization identifiers)

In general, I find the tables more neutral than the decision trees, which inevitable have some subjectivity in them e.g. in the order in which choices are presented (that could lead to other outcomes than when all parameters are presented without hierarchy).

A couple of other comments and questions , focusing on IDs for publications:

  • “journal articles” instead of journals?
  • for books/journal titles, checkmarks for ISBNs/ISSNs in the blue columns
  • is there a specific reason handles have a checkmark for persistent linking and DOIs don’t?
  • Crossref DOIs are specified, but Datacite DOIs are also used for text material (eg. for preprints at Zenodo and Figshare, and in some institutional repositories) (and there are other, sometimes country-specific DOI registries as well)

I also second the comments made above about ORCIDs as identifiers within institutional systems.

Finally, I would be interested whether people working on Wikidata would have any suggestions for additions in the tables.

Looking forward to the next iteration!

1 Like

Thanks so much for all the comments on these guides. We really appreciate the time taken with this considered feedback.

We are starting to work on revising them now and look forward to sharing new versions with you soon. :grinning:

Update: Version 2 of the Guides have been published on Zenodo. We considered all the comments we received in detail but were unable to incorporate all of them into the revised versions. Thanks to everyone who contributed such valuable feedback.


Thanks for this useful guide. It is so good to have in one place the main identifiers covering the different options. At the Publications Office of the EU we manage registration agencies for the DOI, ISSN and ISBN, using them individually or in combination to identify the output from the EU insitutions. I know that many of our authors will be pleased to have access to such a guide. We often end up assigning all three identifiers if the case so demands. For monographs we use ISBN and DOI (as well as ISSN if it is a monograph in a collection) as the libraries who incorporate our metadata into their catalogues want to have a persistent link and they recognise the DOI as a stamp of persistence. But they also want ISBN and ISSN for their own particular attributes.
Here I am putting my hat of also being a member of the board of ISBN on and will concentrate my remarks on the ISBN as others have already commented on DOI and ISSN. I hope you find these will help you to further refine this key resource. So here goes!
The current definition for ISBN is a little light. Perhaps to bring it up to the same level of detail as the others you could consider something along the lines of:
The International Standard Book Number (ISBN) is a unique international identifier for monographic publications. It identifies a specific publication or edition of a publication published by a specific publisher in a specific format. In principle, any text-based, one-off publication available to the public may be identified by ISBN so all formats – from printed to digital to audio and multimedia versions, etc. – qualify within the scope of the ISBN. In addition, an ISBN may be used to identify a whole, single publication (such as a scholarly monograph), a set of related volumes (such as a book series in a defined number of volumes) or a part of a publication such as a chapter or other section, or an article from a serial – provided that these are made separately available. Publishers apply to the national, regional or local ISBN agency that is responsible in their territory and are assigned ISBNs which they may then allocate to their publications. A free look-up service of publishers and their allocations is available at []. Although developed as an analogue identifier, the ISBN can be made actionable by being expressed within the DOI system as an ISBN-A.
Also you put the money sign next to ISBN. Whilst it is true that some agencies charge a subscription or per-ISBN fee, others do not charge a fee. Furthermore, some services are provided free of charge: the ISBN International Agency offers free access to the Global Register of Publishers, many agencies make their publication metadata available online for free.
Decision tree/table and attributes:
In the decision tree above the definitions (and therefore also in the “use cases” table), it might be worth creating two separate entries for “book titles” and “journal titles” since it is not true that ISBN and ISSN can be used interchangeably for these. Also, both ISBN (if it’s a one-off conference) and ISSN (ongoing conference) as well as DOI can be used to identify conference proceedings. ISBN can also identify journal articles (in the decision tree again only Handle and DOI are listed). It might also be worth adding other use cases such as chapters or sections/parts of publications (if made separately available).
In the “attributes” section, you could make a case that, through ISBN-A (DOI+ISBN), the ISBN can provide persistent linking. It depends precisely what is meant by “access and searching (human and machine)” but you could consider that ISBN meets the criterion through the GRP and various books-in-print databases. Similarly, it depends on the precise meaning of “Interoperate, integrate with other identifiers” but ISBN interoperates in the sense that other identifiers (e.g. ISNI, ISTC) can be included in ISBN metadata and the actionable ISBN (ISBN-A) can, of course, interoperate fully digitally.
Happy to discuss in more detail if useful.

Hi Carol

Thank you for getting in touch with these very thorough and helpful comments. Unfortunately as we have already released our second version of the guides and due to the project timeframes, it will not be possible to change them further. I’m very sorry we won’t be able to incorporate your comments at this point.

Many thanks again