innovation in metadata design, implementation & best practices

DCMI Relation_Type Working Group

DCMI Relation_Type Working Group

Chair: David Bearman
Status: This working group is inactive. This working group has been superceded by the Relation Working Group.
Deactivated: 1998-11-04

The Helsinki meeting of the Dublin core working group agreed that there was a one to one relation between a metadata set and an information resource. Thus, if the metadata we are creating describes a digital object whose source was a photograph which in turn has as its source an original manuscript, there will be three metadata sets (each consisting of as many as all fifteen repeatable Dublin Core elements). Obviously we need to define the nature of the relation between these metadata sets (sometimes these will be in the same "record" and other times they will have been created by different people at different times).

  • Initial work on defining a scheme of values for Relation.Type took place in Helsinki. A group of about half the participants (30) met for an hour to define such types and enumerated many relationships which they initially clustered into five categories of relations. It defined these as creative, mechanical, versions, inclusion and reference, but acknowledged that this was a very preliminary scheme.
  • In informal discussions immediately following the meeting, a slightly more complete model emerged when more examples were introduced:
    • Historical
      Creative (such as translations, performances)
      Mechanical (such as reproductions, format transformations, copies, mirrors)
      Versions (such as published editions, drafts, plans, print states)
    • Part/Whole
      Inclusions (such as physical collections, items, parts)
    • References
      Citations (such as bibliographic citations, acknowledgements)
    • Dependency
      Physical (Software or hardware dependency)
      Ideotypical (specimen to Type-specimen, work to iconographic type)
  • No one expects to develop a typology that will be fixed (all discussion was in terms of an extensible list), but the need for a reasonably complete typology is pressing for implementers. This is a call to all those who wish to contribute grist to that discussion to offer examples of relationships which you feel must be accommodated. We have found that examples are very helpful in describing the relationships since we come from quite different communities. This call will be open until November 15, 1997.
  • We will aim to report formally in time for inclusion in the report of the Helsinki conference, which is expected to be submitted in final form before the end of December.


Scheduled Milestone Editor/Contact
Working Group Draft Report, 1997-12-19 David Bearman