innovation in metadata design, implementation & best practices

DCMI Citations Working Group

Dublin Core Metadata Initiative - Citation Working Group

DC9 Citation Break-out Session 2001-10-23 (9:30-11:30am)


Chair: Ann Apps < ann.apps@man.ac.uk>, University of Manchester, UK

Main Points

The Working Group will:

  • Continue to focus primarily on addressing the needs of citing/identifying journal articles
  • Issue:
    • Proposal for citation qualifier for dc:identifier
    • Draft issued by end of 2001
    • Usage Board approval in 2002
    • Proposal for DCSV scheme for dc:identifier
    • Draft issued by end of 2001
    • Usage Board approval in 2002
    • Need advice on XML & RDF encoding for DCSV
  • Defer proposing OpenURL as a scheme

Report of DC9 DC-Citation Working Group Session

1. Overview of purpose of group, previous work and current status of proposals.

  • The WG is initially focussing primarily on the bibliographic citation metadata for a resource itself, not yet on linking metadata. The WG has restricted itself for now, in its current charter, to journal articles.
  • This issue has been discussed at previous DC Workshops, but recommendations made by the previous WG were not taken forward by DCMI. The WG has been re-opened because it is obvious that recommendations on encoding bibliographic citation information are still needed.
  • Journal article citation information includes identification (mostly optional and some, but not many, repeatable) of:
    • Journal: title, abbreviated title, identifier (ISSN)
    • Issue: volume, issue/part number, chronology, (identifier eg SICI)
    • Article: page range (other article information would be captured in usual DC elements)
  • The WG has decided that this information should be captured in dc:identifer, which recognises that the citation information identifies the article. This decision was informally endorsed by voting in the final Plenary session of DC8.
  • The current proposal, 'Recording Bibliographic Citation Information for Journal Articles in Dublin Core' (a copy was included in DC9 attendees' information pack), includes in one document:
    • a 'citation' qualifier for dc:identifier
    • a DCSV scheme, working name DCCITE, for citations
    • how to encode a citation in OpenURL (web transportable metadata)
    • an XML and RDF syntax for encoding a citation
    • ISSN should be encoded in dc:relation with an isPartOf qualifier
    • a suggestion that dc:relation and dc:source could have a citation qualifier
  • This current draft proposal needs readdressing, and splitting, because some of its recommendations are not consistent with the current DC model of elements and terms (qualifiers), and there is too much in one document.

2. A 'citation' qualifier (element refinement) for dc:identifier, why is this needed?

  • It could be implied by an encoding scheme, eg a DCSV-DCCITE, OpenURL.
  • A majority voted for this qualifier at DC8.
  • It would allow a free text citation, or a citation according to a publisher's standard text format.

3. A DCSV , working name DCCITE.

  • This would contain: journalTitle, journalAbbreviatedTitle, journalIdentifier, journalVolume, journalIssueNumber, journalIssueDate, pagination.
  • These 'sub-elements' need definitions of their meanings.
  • We need a name for the scheme if we decide to go ahead with this part of the proposal.
  • Do we need to define ordering of sub-elements or which are repeatable, mandatory?
  • journalIssueDate is the 'cover date', not the 'issued' date - this would need definition in any proposal document. It would probably not have a defined encoding for flexiblity in recording eg seasons.

4. OpenURL.

  • Really for linking data, not resource metadata
  • Proposed NISO standard (on fast track) but not yet a standard, so may not yet be stable. Some things needed for resource metadata eg. abbreviated title, may be dropped.
  • However, OpenURL will be versioned and the initial draft will be fixed as version 0.1 (this has now happened).
  • Probably not able to capture repeatable sub-elements.
  • Not really human readable.
  • An OpenURL, according to its syntax, should include a Base URL. We are thinking of using partial OpenURLs (ie without the Base URL). There is a question, as yet unanswered, about whether a partial OpenURL could/should be considered a standard OpenURL.
  • An OpenURL can include information encoded elsewhere in a DC record, eg article title.
  • It is likely that applications providing OpenURL linking would generate the OpenURLs on-the-fly from citation sub-elements recorded separately, rather than holding complete OpenURLs.
  • Is this an either/or with DCCITE, or would we propose both?

5. Citation qualifier for dc:relation and dc:source

  • This is effectively using the citation as linking data, so an OpenURL would be appropriate here as an encoding scheme.
  • dc:relation already has qualifiers, eg references, isReferencedBy, isPartOf, so doesn't really need a citation qualifier.
  • This is something to look at later.

6. XML and RDF encoding of DCCITE sub-elements

  • Encoding of the 'sub-elements' of a DCSV as XML elements is going beyond current DC usage recommendations.
  • Other DCSV schemes, eg DCMI Period, include XML syntax suggestions but with a disclaimer that this isn't yet endorsed DCMI usage.
  • We need to ask the Usage board for advice on this. We should write a draft proposal, then ask them for advice, before writing a formal proposal.
  • Possibly this should be defined within a 'citation' profile or namespace.
  • Requirement for XML/RDF is significant for resource description.

7. Encoding schemes

  • DCMI usage board will be making avaliable a means to register standard encoding schemes.
  • Citation WG will need to register appropriate schemes, eg ISSN, SICI, possibly OpenURL.
  • Should these schemes be part of a 'citation scheme' (profile) or just dc:identifer schemes?

8. Consensus
The meeting agreed on several significant points:

  • A citation qualifier for dc:identifier is required.
  • A citation should be human readable.
  • Work on producing a formal proposal and recommendation for journal article citations should be done as quickly as possible - lots of groups/projects need this information.

9. Other discussion and points raised

  • Citations for books and chapters could use BICI.
  • journalIssueDate encoding could be of interest to newspaper publishers.
  • Abbreviated journal titles are problematic. There are several schemes, and many references at the end of journal articles do not adhere to any of these.
  • Some e-journal articles do not have pages: use a standard citation without pages?; introduce another article identifier? We should re-examine this later.
  • What is a journal? Maybe something which has an ISSN.

10. Open Issues for later work
These open issues for the WG were noted briefly:

  • Other genre
  • Journal (and genre) specific resource citations (liaise with Type WG).
  • Author affiliation (liaise with Agent WG). Note that an affiliation on a scholarly work persists even when an author moves.
  • Linking data.
  • Citations for electronic-only journal article.

11. Work Plan
Note that this is dependent on agreement by the Citation WG on decisions made at the meeting.

  • Proposal for a citation qualifier for dc:identifier
    • Draft by end 2001
    • Formal proposal for Usage Board approval in 2002
  • Proposal for a DCSV scheme, working title DCCITE, for dc:identifier
    • Draft by end 2001
    • Formal proposal for Usage Board approval in 2002
  • Proposal for XML and RDF encoding for DCSV
    • Request advice from Usage Board
  • Defer proposing OpenURL as a scheme.

2001-12-07