Workbook on Digital Private Papers > Administrative and preservation metadata > Persistent identifiers
Persistent identifiers
What do we need to identify?
This understanding of what a repository might need to identify is partly based on the use of identifiers in the PREMIS Data Dictionary for preservation metadata.
The archive The archive, as a collection, must have an identity so that it can be associated with objects belonging to it using their identifiers. archive:1 archive:2 |
An accession In order that digital manuscripts can be associated with an accession, an accession id is needed. archive: accession:1 accession:2 |
A digital manuscript (the intellectual entity) This is the conceptual item that might be described in a catalogue, and will be given a traditional identifier, such as a reference code or shelfmark, when it is catalogued. An example is: archive:1 accession:1 object:1 - Website of Politician X, 20 Jul. 2006 (also known as MS. Eng. 23) |
A representation A representation is a particular instance of an intellectual entity, so MS. Eng. 23 above could have two representations as follows: archive:1 accession:1 object:1 [MS. Eng. 23] representation:1 - Website of Politician X, 20 Jul. 2006 (html and jpeg files) representation:2 - Website of Politician X, 20 Jul. 2006 (PDF capture of website) |
|
A file A representation is composed of one or more files, for example: archive:1 accession:1 object:1 [MS. Eng. 23] representation:1 - Website of Politician X, 20 Jul. 2006 file:2 - html file file:3 - jpeg file representation:2 - Website of Politician X, 20 Jul. 2006 file:5 - pdf file |
|
Different representations will be created for:
- The ‘original’ digital object which was deposited with the archive.
- Successive migrations of the data object deposited with the archive.
A repository’s preservation strategy may require that digital objects be migrated to other formats in response to format obsolescence, or on ingest if the repository has limited its support to a small number of formats. The migrated objects will need their own PID; this PID can be used to relate the migrated object to the object that it was derived from and also to the traditional identifier in the archival catalogue which describes the digital manuscript.
Metadata (and versions of metadata) describing the object
Metadata describing objects may also need identifiers; this enables the repository to associate one piece of metadata to one or more objects. Metadata which may need an identifier includes:
| Event metadata: e.g. event:1 | Agent metadata: e.g. agent:1 |
| Rights metadata: e.g. rights:1 | Descriptive metadata: e.g. catalogue:1 |
