Meaning and purpose
An identifier is a sequence of characters or words that uniquely identify an object within a particular context or the domain of a specified authority. It may be globally unique if the specified authority has a global domain.
Identifiers that are useful for discovery, linking of metadata records, re-use, or to support the functions of the metadata provider, can be recorded, even if they are not unique or persistent.
Use the following suggested vocabularies or other standards:
- abn: Australian Business Number
- arc: Australian Research Council identifier--identifiers issued by or on behalf of the Australian Research Council
- ark: ARK Archival Resource Key, a persistent identifier scheme from the California Digital Library at the University of California
- AU-ANL:PEAU:National Library of Australia party identifier(formerly nla:)
- doi: Digital Object Identifier Name
- handle: HANDLE System Identifier
- infouri: 'info' URI scheme--administered by OCLC on behalf of NISO
- isil: International Standard Identifier for Libraries and Related Organizations
- local: identifier unique within a local context
- orcid: Open Researcher and Contributor ID
- nhmrc: National Health and Medical Research Council identifier
- purl: Persistent Uniform Resource Locator see also National Library of Australia use of purl
- uri: Uniform Resource Identifier--a compact sequence of characters that identifies an abstract or physical resource
Research domain examples
Identifiers for research data collections may be very context-specific, as these examples from a record in the Tardis Federated Diffraction Image Publication Repository show.
- experiment/view/11 (type local)
- 3BJX (type pdb (Protein Data Bank identifier))
While valuable in a local context, as well as useful for discovery, these identifiers are not suitable for use as keys in the ANDS Collections Registry.
Use in Research Data Australia
Identifiers are displayed and searchable. Where possible, identifiers will be displayed as resolvable URLs, see information about URL displays.
Where two or more records, from the same or different data sources, share a common identifier, the records will be treated as describing the same thing. In Research Data Australia, these records are merged into a single search result and links to each of the merged records are displayed on the view page of each record. See Step 8 of Best practice for creating party records for details.
RIF-CS best practice guidelines
For all records
Identifiers for collections, parties, activities or services being described, including non-unique and local identifiers, should be recorded in the identifier element. The <identifier> element can be repeated if multiple identifiers need to be described.
NOTE: Do not use the identifier for an object (e.g. a dataset's DOI, a party's NLA identifier) as the key for a metadata record describing that object—the metadata record needs its own unique separate identifier.
Best practice for collection identifiers
It is recommended that all identifiers be persistent i.e. of type 'doi', 'handle' or 'purl', but any URI is acceptable. The dataset is not required to be online but there must be a resolvable URI at a minimum.
For detailed information on persistent identifiers, see the ANDS Guides on Persistent Identifiers (awareness, working and expert levels) and Digital Object Identifiers, and the ANDS Guide to Identify My Data.
Best practice for activity identifiers
Party records should record the NLA party identifier in the RIF-CS Identifier element.
This will allow linking of party records to party records supplied to the ANDS Collections Registry by the National Library of Australia. Those NLA party records will use NLA party identifiers as keys. In the screenshot below, the NLA persistent identifier for Margaret Joy Kartomi is http://nla.gov.au/nla.party-493210 , and this will be used as the key for the NLA-supplied party record in the ANDS Collections Registry. Only the NLA should use this persistent identifier as a key.
(NOTE: The NLA party identifier should NOT be used as a RIF-CS key for any records except party records from the NLA.)
Trove People and Organisations results screen showing persistent identifier
Obtaining party identifiers
The ARDC Party Infrastructure project will allow party records to be contributed to Trove and be allocated NLA party identifiers. More information
Recording party identifiers
If the NLA Party Identifier is known for a party then it should be supplied in an identifier element. AU-ANL:PEAU is an internationally used ISIL code for the National Library of Australia's authoritative list of names for Australian people and organisations and should be used as the identifier type. While the URI form is preferred, NLA Party Identifers may be represented in either of the forms shown in these examples.
If an institution supports its own URI identifier scheme for researchers and/or staff then they should supply this identifier in their party record in the Identifier element.
Publisher identifier or other international identifiers
If other researcher identifiers are recorded in institutional systems, for example, the ResearcherID from ThomsonReuters, or ORCID (Open Researcher & Contributor ID), then these should also be provided as they will provide additional linking information. Other international identifier systems are being developed and are likely to be interconnected in the future, for example VIAF (Virtual International Authority File). If these are available, including them in the party record's identifier element will assist both discovery and record linking.
If the identifier is available as a stable URL, provide it using the URI type so it displays as a resolvable link in Research Data Australia.
If the identifier is not available as a resolvable URL, provide it using a type that reflects the authority for the identifier. This will display as text only (not a resolvable link) in Research Data Australia. While not currently included in the ANDS suggested vocabulary, ANDS recommends using the types below.
If an identifier is usually expressed in the form of a query (for example, http://www.scopus.com/authid/detail.url?origin=resultslist&authorId=15054803200&zone= , also use these authority-based types.
Identifiers for collections
Identifiers for parties
Identifiers for activities
Australian Research Council identifier example:
NHMRC identifier example:
|April 2010||Consultation draft|
|26 Oct 2010||Changed guidelines for use of National Library of Australia identifiers|
|14 Oct 2011||Removed 'txt' suggested identifier type (previously included in error) — use 'local' instead of this type|
|14 Oct 2011||Added link to best practice for activity identifiers, clarified types to use for party identifiers (ThomsonReuters, Scopus, VIAF, ORCID), added link to Contributing to Trove page
|29 Nov 2011||Added link to information about URL displays for identifiers|
|13 Mar 2013||Added further information about ORCID identfiers
|26 Nov 2013||Added nhmrc as a new identifier type|
|27b Mar 2014||Added information to Use in Research Data Australia section|
Please send any feedback on this page to firstname.lastname@example.org