Lookup Release by Identifier


Digital object identifier: "it's not an identifier for a digital object, it's a digital identifier for an object". Except they are pretty much all digital objects. Fatcat doesn't include all DOIs (eg, for granular components or TV shows), but it should for all complete research publications. DOIs are generated by publishers, who pay a fee to registrars like Crossref, Datacite, or JALC.

  Don't include a URL like "http://dx.doi.org/"


PubMed/MEDLINE is a catalog of biomedical publications. Catalog updates are published frequently, and may include deletions. Fatcat currently is only updated with new entries, it does not delete PMIDs. PubMed is edited by a large staff of trained catalogers, and has strict inclusion criteria for journals.

PubMed Central Identifiers (PMCID) reference works deposited in the US (or European) PMC repositories. Not all works with a PMCID have a PMID. PMCIDs can be versioned (eg, with a trailing ".2"), but these are rare and most Fatcat releases only have the non-versioned PMCID.

  Include the "PMC" prefix.

Wikidata QID

Wikidata is the structured, machine-readable database that complements Wikipedia. Entities, which can include bibliographic entities like books and papers, are identified by a "QID". Fatcat also links container (journal) and creator (author) entities to Wikidata by QID. Anybody can edit Wikidata and create QIDs.

  Include the "Q" prefix.


Archival Resource Keys are no-cost, distributed identifiers based on URLs. Mostly used by archival systems (as opposed to publishers or repositories).

  Only the identifier part (starting with "ark:/"), not the full URL.


International Standard Book Number, assigned to specific published versions of a book (including ebooks).

  Canonical ISBN-13 number (not ISBN-9), with hyphens


arXiv identifiers are generated for works deposited in the free arxiv.org pre-print repository. Identifiers are generated for each version of a work, eg ending in "v3". Fatcat releases must be versioned; if you don't know the version you can always lookup "v1" and then look for other releases under the same work.

  Old and new style supported, but version always required for exact lookups.


Handle identifiers are registered under the hierarchal handle.net system, maintained by CNRI. While technically all DOIs are Handles, our system treats them as distinct identifiers.

  Only include the identifier itself


JSTOR identifiers are numbers internally assigned to works archived on jstor.org. Many works also have a DOI with the number after a 10.2307/ prefix, but not all works have such a DOI. Fatcat mostly has JSTOR identifiers for works in the "Early Journal Collection", which are in the public domain.