The first version of our metadata input schema (a DTD, to be specific) was created in 1999 to capture basic bibliographic information and facilitate matching DOIs to citations. Over the past 20 years the bibliographic metadata we collect has deepened, and we’ve expanded our schema to include funding information, license, updates, relations, and other metadata. Our schema isn’t as venerable as a MARC record or as comprehensive as JATS, but it’s served us well. It’s not currently positioned to fully support everything we want to do long term - we’d like to support assertions, map cleanly to JATS and schema.org magically at the same time, and maybe even move beyond XML - but for now it’s something we can work with to empower member metadata to help find, cite, and connect scholarly content.
We’ve maintained backwards compatibility for most things since 2007 but this update will require some moderate changes to how contributors are modeled. The balance between supporting established tagging and addressing the evolution of what we collect and how it is expressed can be tricky. We want to collect good metadata without significantly disrupting the workflow of our membership, who are the source of the metadata. Even so, this is a fairly pragmatic update that will position us well for the future. I look forward to supporting new types of content and metadata in the future, but for now take a look at what I’m proposing.
I’m proposing some updates and additions to the metadata we collect, and would like your feedback. To fully and elegantly support affiliation identifiers and multiple author roles, we need to break backwards compatibility. Specifically, we want to:
The CASRAI CRediT taxonomy is increasingly used to represent roles common to contributors to research outputs. Our members are applying CRediT to contributors, so we want to capture them as well. Supporting CRediT allows Crossref and our membership to identify and credit contributors beyond authors and editors.
As most of you know, a contributor often does more than one thing - they write, they edit, they curate. We currently only allow one contributor role as an attribute, but, to realistically support CRediT and accurately capture evidence about the work, we need to allow multiple contributor roles. This will break backwards compatibility. We can potentially support the old way and the new way, but I’m trying to avoid awkward compromises wherever possible.
Supporting CRediT doesn’t mean you need to adopt CRediT. We’ll continue to support existing author roles, but they’ll be marked up differently. Details are in our request for feedback document.
We collect ORCID iDs in our metadata but do not currently support other types of contributor identifiers. We also don’t support affiliation or organization identifiers beyond those assigned within our funder and clinical trial registries. We’ve had increasing demands from both metadata suppliers and users to expand support for affiliation identifiers because…identifiers are useful. We also want to expand author identifier support as ORCID IDs may only be registered by researchers who are able to curate their own ORCID record. Adding support for ISNI and Wikidata IDs is a common request, but we anticipate there’s a need for other identifiers as well.
Our plan is to accept identifiers registered with identifiers.org as well as other identifiers upon request. We prefer to remain consistent with the identifiers.org registry as much as possible.
We’re particularly keen to support open community-led identifiers like ORCID and ROR and will continue to do so, but also want to support the metadata our members want to distribute. Organization identifiers will be particularly useful as they’ll help us populate records with ROR IDs in the future, leading to better quality affiliation metadata.
We currently require a surname for all contributors, and don’t provide comprehensive support for contributors whose names are represented by multiple alphabets, or who have nicknames or aliases, or who don’t have a surname. To begin with, we’ll replace surname with the more widely used ‘family name’ and remove the fixed surname requirement, allowing only a given name to be provided where appropriate. We’ll also allow a variety of names to be provided for each contributor.
We currently collect affiliation as a single string - we’re going to break that up to support affiliation names, and add in support for organizational identifiers like ROR.
For those of you who send us references, we’re adding a few fields to better support data citation. We’re also going to allow you to (optionally) supply a specific publication type for references.
We’re making some other small updates as well. If you have a small request, we may be able to accommodate it in our next update. Larger changes or additions will probably have to wait for future updates, but we’d love to start collecting suggestions now.
I’ll be giving a webinar on December 19 at 02:00 and 15:00 UTC to go over these changes in detail - please visit our webinars page to register.
Again, please leave feedback, ask questions, and make suggestions in the feedback document, or if you prefer send feedback via email to feedback@crossref.org. We’ll be taking feedback through January 15, 2020.
Version 5.4.0 metadata schema update now available
2025 March 19
2025 public data file now available
2025 March 12
