7. Identification of Metadata Formats for Further Consideration
A detailed review of metadata formats was undertaken as part of the DESIRE project and is now available as a deliverable (draft). We do not intend to repeat the detailed examination of the full range formats contained in that report. Here we shall attempt to summarise characteristics of those formats which might be used as solutions within the context of BIBLINK. We will also include consideration of publishers' formats which were not included in the DESIRE overview.
The various categories of players involved in the information flow are associated with different metadata formats as described in section 6 above. From this list we have chosen particular formats as suitable for the requirements of both publishers and national bibliographic agencies, and as likely options for use within BIBLINK.
The criteria for selection are:
The formats chosen for further investigation are:
- familiarity of use within the relevant sectors
(traditional publishers, web publishers, national libraries, booksellers and other bibliographic agencies)
- appropriate level of detail
(note that according to the present scoping of the project it is unclear what level of detail is required. This will be dependent on such factors as whether CIP records are required, and whether the resource being described originates from traditional publishing or web publishing. For this reason formats at different levels of detail are included)
- ability of the format to accommodate description of electronic resources
(we do not intend to consider formats where there is no signal of adoption of elements appropriate for electronic resource description)
- the format has some grounding as an international standard
(There is an intention among the stakeholders to achieve acceptance as an international standard in whatever forum is appropriate)
- the cost of creating the record is appropriate to the environment
(It is assumed that detailed descriptive records, as categorised in Band 4 section 5.1, will not be created by publishers or national libraries. However we will assume that the information in such records may well feed into the process. So publishers such as archive services who create records in CIMI, IPCSD or EAD format may wish to map these records into another preferred format to feed them into the process of bibliographic control).
Simple Internet descriptions:
- Dublin Core.
- MAJOUR: Modular Application for Journals, European Workgroup for SGML, 1991
- SSSH: Simplified SGML for Serials Headers, London, Book Industry Communication and PIRA International, 1996
- EDIFACT formats for book and serial product information.
- TEI (and its relation to CIMI, EAD).