BIBLINK - LB 4034 - D4.1 Format Conversion Feasibility

Format Conversion Feasibility
Work Package 4 of Telematics for Libraries project BIBLINK (LB 4034)

10. Mapping National Libraries' metadata requirements to Dublin Core.

10.1 Introduction

To assist consensus building the CIP data elements agreed by the national libraries are here mapped to Dublin Core. The mapping tables in this report should inform further discussion on content and usage of each element and help to identify where qualifiers and extensions would have to be made to the Dublin Core elements. Further discussion of this mapping will need to involve publishers as regards what data they can provide. Final decisions will need to be made regarding which DC qualifiers will be used, and what extensions will be added to the 'base list' of DC elements. It is important that issues arising from the mapping of DC to UNIMARC are also taken into account when defining the data elements required from publishers.

D1.1 suggests that national libraries wish to create records in various flavours of MARC and would intend to apply detailed cataloguing rules to the content of the records. The formats used by the participating libraries vary, but are usually based on ISBD or AACR2. For the purpose of this report, it is assumed that ISBD or AACR2 style formats would be desired for the national libraries' metadata requirements for electronic publications.

10.2 DC Qualifiers

The DC-4 meeting in Canberra (March 1997) proposed the formal identification of the structure of elements and possible qualifiers in Dublin Core [15]. In response, Rebecca Guenther has recently produced a proposal for Dublin Core qualifiers/substructure, which includes specific proposals for the qualifiers "scheme" and "type" [16]. These proposals are currently under discussion in the Dublin Core community. Guenther reiterates the Canberra meeting's insistence that the "type" qualifier should only be used to refine elements, not to extend their semantics and that each element should have a default meaning. The qualifiers can be understood as follows:

SCHEME interprets the value in the content and is generally based on external standards.

TYPE refines the definition of the data element itself.

If SCHEME and TYPE can not meet these principles, then an extensibility mechanism should be used.

10.3 Extensibility

Where the national libraries' metadata requirements cannot be described using Dublin Core - with or without qualifiers - then new elements can be proposed.

10.4 The mapping

BIBLINK Data Element	Dublin Core
Author	Creator If a distinction needs to be made between personal and corporate authors, DC can differentiate between Creator.Personal and Creator.Corporate. Additionally, the SCHEME "Library of Congress Name Authority File" can be used if appropriate.
Contributor	Contributor As with the Creator element, distinctions between Contributor.Personal and Contributor.Corporate can be made, as can a SCHEME for the Library of Congress Name Authority File.
Date of publication	Date The proposed DC SCHEME default for Date is ISO 8601 with six proposed levels of granularity: Year: YYYY (e.g. 1997) Year and month: YYYY-MM (e.g. 1997-07) Complete date: YYYY-MM-DD (e.g. 1997-07-16) Complete date plus hours and minutes: YYYY-MM-DDThh:mmTZD (e.g. 1997-07-16T19:20+01:00) Complete date plus hours, minutes and seconds: YYYY-MM-DDThh:mm:ssTZD (e.g. 1997-07-16T19:20:30+01:00) Complete date plus hours, minutes, seconds and decimal fractions of a second: YYYY-MM-DDThh:mm:ss.sTZD (e.g. 1997-07-16T19:20:30.45+01:00) It is unlikely that more than the first three levels would be relevant in the context of BIBLINK.
Description	Description
Edition/version	Extension to DC required.
Extent (size)	Extension to DC required.
Format	Format In DC terms, Format refers to the data representation of the resource, including things like Postscript or text/html. There is still some debate concerning the desirability of using enumerated lists of format types but the default as currently proposed is free text.
Frequency	Extension to DC required.
Hash Value	Extension to DC required.
Identifier	Identifier It is proposed that URL is the DC default identifier. Therefore any other identifiers used, e.g. ISBNs, ISSNs or DOIs, will have to include a SCHEME in the DC record.
Keywords	Subject Keyword is the default for DC Subject.
Language	Language It has been suggested that the content of this is DC should coincide with NISO Z39.53 three character codes, although the default scheme is free text. If USMARC/Library of Congress style language codes are used (e.g. as in UNIMARC), the SCHEME given should be "Z39.53".
Place of publication	Extension to DC required.
Price	Extension to DC required.
Publisher	Publisher
System requirements	Extension to DC required.
Terms and conditions	Rights The default for Rights in DC is free text, although it is intended for a link to an URL.
Title	Title

10.5Notes

In the BIBLINK context, there may be a requirement to extend DC metadata to include the following elements:

Edition/Version

Extent

Frequency of Publication

Place of Publication

Price

System Requirements

Some interaction with name authority files for DC Creator and Contributor would be useful or desirable. Guenther already has a DC SCHEME for the Library of Congress Name Authority File, but additional SCHEMES might be required for partner libraries. Thought could also be given to the inclusion of particular subject schemes rather than just keywords, then LCSH or classification numbers, for example, could be included

If an ultimate mapping to UNIMARC or other MARC formats is desirable, it may be worth trying to maintain differences between personal and corporate in both Author and Contributor fields.

10.6 Some examples

Please note that the following examples are not intended to be definitive.

10.6.1 Web Page:

Metadata for a Web page, when translated into the BIBLINK data elements (incorporating Dublin Core), might look like the following:

Author (corporate)	DC.creator.corporate: Cambridge University Library
Title	DC.title: Taylor-Schechter Unit Home Page
Date	DC.date: 19970605
Language	DC.language: eng
Format	DC.format: text.html
Keywords	DC.subject: Taylor-Schechter Genizah Research Unit; Cairo Geniza, papyrus
Identifier	DC.identifier: http://www.lib.cam.ac.uk/Taylor-Schechter/
Place of publication	Cambridge
Publisher	DC.publisher: University of Cambridge

10.6.2 CD-ROM:

In comparison, a commercially published CD-ROM could be described using the BIBLINK data elements (incorporating Dublin Core) in the following way:

Author (personal)	DC.author.personal SCHEME=Library of Congress Name Authority File: Migne, J.P. (Jacques Paul), 1800-1875
Title	DC.title: Patrologia Latina Database
Date	DC.date: 1993
Language	DC.language: lat
Format	CD-ROM
Extent	2 computer laser optical disks ; 4 3/4 in
Description	DC.description: The Patrologia Latina Database is an electronic version of the 221 volumes of the first edition of Jacques-Paul Migne's Patrologia Latina which was published between 1844 and1865. The Patrologia Latina comprises the works of the Church Fathers from Tertullian in 200 AD to the death of Pope Innocent III in 1216. The database is fully searchable.
System requirements	Multimedia PC 486x or higher, 8mb memory, CD-ROM drive, sound card, SVGA 256-colour monitor, Windows 95 or Windows 3.1
Keywords	DC.subject: Early Christian Literature; Patristics; DC.subject SCHEME=LCSH: Christian literature, Early -- Latin authors – Texts DC.subject SCHEME=LCSH: Fathers of the church, Latin -- Texts
Identifier	DC.identifier SCHEME=ISBN: 0-89887-113-1
Place of publication	Cambridge
Publisher	DC.publisher: Chadwyck-Healey

References

[15.] Weibel, S., Ianella, R. and Cathro, W. The 4th Dublin Core Metadata Workshop report. D-Lib Magazine, June 1997,
<URL:http://www.dlib.org/dlib/june97/metadata/06weibel.html>

[16.] Guenther, R. Dublin Core qualifiers/substructure: a proposal 15 April 1997. <URL:http://www.loc.gov/marc/dcqualif.html>