A start on some candidate vocabulary and testing a vocabulary tool

27 Nov 2013


We are testing out some candidate vocabulary using RDA Term Collection Wiki prototype.
The small attachment to this positng  illustrates this using the Wittenburg model sketch which has a series of related assertions that include key a functional grouping of terminology. The idea is to use these diagrams to evoke the terminology needed to understand them.  The take the key assertion of the diagram to be something  like this:

1.    A PID record that points to a metadata record and to instantiations of identical bit-streams that may store additional attributes (in a repository);
I     I have some of the terms underlined and these now have candidate definitions which have been placed in the tool to encourage its testing as well as discussion of the candidate definitions using the tool.  Some example are provided in the attachment and reproduced here.  For example:
1.    A persistent identifier (PID) is generally agreed is a long-lasting (non-corrupting) reference to informational content that distinguishes a digital or data object (such as a document, an individual file or a collection of files) from other DOs.
2.     A Data Object is a type of object or collection composed of discrete data, considered as a unit, representing an instance or instances of a datatype that is known or assumed to be known [adapted from ISO/IEC 2382-17]. Data objects are referenced by identifiers (IDs). A data object or record’s content may be expressed at bitstreams and there may be multiple instances of the same bitstream called a copy of a data object.


File Attachment: 
  • Yde de Jong's picture

    Author: Yde de Jong

    Date: 28 Nov, 2013

    Dear Gary,
    Thanks for this!
    I see quite some similarities with the Knowledge Organization System (KOS) for biodiversity information, especially the Term Forum WiKi.
    Did you communicate with the GBIF people on this (especially Eamonn O Tuama)?
    Kind regards,
    Op 27 nov 2013, om 22:21 heeft Gary het volgende geschreven:

  • Andrew Treloar's picture

    Author: Andrew Treloar

    Date: 13 Dec, 2013

    Gary (and others),
    thanks for the open process you are following to engage people with the activities of the group.
    In that spirit, I’d like to comment on the definition that you give below for persistent identifier.
    This describes it (I’m paraphrasing the definition now) as pointing to metadata about the object rather to the object itself. This is certainly a way in which PIDs can be used, but it is not the only way. There is nothing stopping a PID resolving directly to the object itself. In fact, many DOIs point directly to journal articles, rather than to the article metadata/landing page. I think it might be good to nuance the definition a bit.
    The definition also doesn’t talk about the distinction between the identifier itself, and the resolution mechanism. Both of these need to persist and not degrade for the system to function as intended. In fact, one can identify five different persistences that are required:

submit a comment