Explanation
How do I fill this in?
Repeatable:
Yes
While multiple local identifiers may be entered, it is recommended to make sure at least one identifier is a unique, system-generated value.
Recommended Data Entry Type:
Free text or URI
Locally generated system-generated ID numbers or Persistent IDs (PIDs) are recommended when feasible, as opposed to filenames or call numbers/accession numbers. Filenames may be used as identifiers, but they are recommended as additional/secondary identifiers because they may not always be unique.
For digital assets that don’t have a readily available system ID number, enter the URL or filename of the digital resource. In cases when a URL is the only viable identifier, the Access URL element may use the same value as Identifier – Local.
Note that there are separate Core metadata elements for recording the call number or accession number that corresponds to an original source object (see Source Item Call Number), and for using a standard/global identifier (see Identifier – Standard).
Punctuation/style notes:
When recording a URL as an identifier, make sure to preserve the URL’s original capitalization and punctuation: even minor changes to file and folder names in a web address can result in broken links.
Additional detail:
If your encoding standard or authoring tool supports it, it is helpful to specify which type or convention of local identifier you are recording. If the identifier was generated by a specific system, it is also helpful to indicate which system the identifier has been generated by.
System Tips:
When migrating metadata from one system to another, it is very helpful to retain original/previous systems’ IDs in order to troubleshoot data migration issues.
When providing IDs for records to be ingested into Primo, system-generated IDs are strongly preferred as opposed to filenames, URLs, or other conventions. When mapping to simple Dublin Core for Primo, it is important to either designate which is the local system ID (if multiple identifiers are present) or to only map the preferred, Primo-friendly local identifier to the Dublin Core identifier element.
Dataverse generates a Handle URL for data sets as they are deposited, which is used as the local system ID. Individual files associated with the data set receive local system-generated ID numbers.
Emory’s Fedora repository framework utilizes an Emory-developed PID Manager service, which generates ids in ARK format for all assets ingested into the system.
Help/Troubleshooting
What are best practices for good filenames?
Filenames are especially important when viewing a digital object in isolation without accompanying metadata. It is recommended to consider a filenaming scheme prior to creating/digitizing content. For websites, filenames can impact search engine optimization. For other types of digital content, filenames may follow either human-readable or machine readable conventions. The Technical Guidelines for Digitizating Cultural Heritage Materials (FADGI) provides guidelines for constructing filenaming schemes (see page 79). Note that filenames may lose their meaning and context when files are moved from one system environment to another.
How do I request PIDs through Emory’s PID service?
Emory’s PID service can be utilized by submitting a Service Now ticket, and referencing Emory’s PID service (pid.emory.edu). You can request that your identifier be formatted as a PURL or an ARK.
My item has multiple identifiers generated by different systems - which one should I use?
While it is helpful to retain legacy system identifiers, if your item has been migrated across several systems, prioritize the identifier from the system which is most actively used (for example, the ID from the system where the metadata is most likely to be edited).
FAQs:
What is a URI? What's the difference between a URI, URL, and URN?
Links and Resources:
DLF Best Practices for Shareable Metadata – Identifying the Resource