Crossref metadata: what a journal's records should carry
Crossref's Participation Report shows, in public, which elements a journal's records carry. Each element does something for the journal; each blank costs it something. What they are, and how the blanks are filled from what the journal already has.
References
Crossref's Cited-by counts a citation only when the citing article's record carries its reference list, matched to DOIs. A journal whose records have no references gives its own authors no citation counts, and gives none to the journals it cites.
Abstracts
OpenAlex, the discovery services and the AI assistants read the abstract from the record. Without it an article is a title.
ORCID iDs
With the author's iD on the record, ORCID adds the article to the author's profile and the author's institution sees it. A name alone matches nobody.
Affiliations and ROR IDs
The institution named as the record's own field, with its ROR ID, is how the institution's output is counted — by OpenAlex, by funders, by the rankings that read them.
Funder IDs and award numbers
The funder's registry ID and the grant number on the record are what a funder reads to find the work it paid for; an acknowledgement in the PDF is read by no one.
Licence URLs
DOAJ and Unpaywall read the licence from the record. "Open access" on a web page is a claim; a licence URL on the record is a fact.
Crossmark
For journals with a Crossmark policy: corrections, retractions and expressions of concern declared on the record and shown on the article, so a reader of a retracted paper knows.
Full-text links
The links for text and data mining and for Similarity Check, so the journal's own articles are compared against new manuscripts and read by the tools that build on them.
Read your journal's report, and what a repair would change
DOI Doctor reads every record of the journal back from Crossref, shows the report as it reads today and as it would read after the pending repair, element by element — and sends nothing until the journal presses Approve.
Questions people ask
What is Crossref's Participation Report?
A public page for every Crossref member showing, element by element, what share of its records carry references, abstracts, ORCID iDs, ROR IDs, funder information, licences, Crossmark, Similarity Check links and text-mining links — for current articles and for the backfile separately. Anyone can open it; the indexes and the funders do.
Why does Cited-by show nothing for our journal?
Almost always because the records carry no reference lists. Cited-by is computed from the references deposited with each record; a journal that deposits none receives no counts and gives none. Depositing the references of the existing articles is the repair, and it is done without re-registering a single DOI.
Can the metadata of DOIs already registered be completed?
Yes. A record is updated by depositing it again in full with the missing elements filled; the DOI does not change and nothing is charged for the update. DOI Doctor reads the missing elements from the article itself — the PDF, the article page, the JATS or NLM file — shows the journal what it found, and deposits after approval.
Where do abstracts and references come from?
From what the journal already has. The abstract is read from the article; the references from the reference list, each matched to its DOI where one exists; ORCID iDs and affiliations from what the paper prints; the licence from the journal's policy. The journal approves each line before it is deposited.
