The Dublin Core Metadata Initiative

by Susan Haigh
Network Notes #63
ISSN 1201-4338
Information Technology Services
National Library of Canada

December 1999

1.0 Introduction
The Dublin Core metadata scheme comprises 15 data elements for describing information resources to support resource discovery in Web-accessible applications. This element set emerged from a series of international invitational workshops that have been held since 1995, at which broad consensus was achieved among experts in resource description, networking, encoding standards, information retrieval and a range of subject disciplines. The Dublin Core Metadata Initiative is the open body that continues to shape, refine and promote the standard.

Owing to its international support and its rapid acceptance as a means of resource description by many communities, Dublin Core has emerged as the most important standard for the simple description of electronic information resources. Significantly lighter than MARC, Dublin Core is likely to become important to libraries as an alternative to full bibliographic cataloguing for Internet-based resources. While not likely to supplant MARC for the bibliographic description of resources where the use of MARC is in place and sustainable, Dublin Core is the leading alternative resource description standard for applications such as subject gateways and diverse digital collections. It also represents a foundation for a broadly interoperable semantic network based upon a basic element set that can be widely used. The Dublin Core standard does not preclude the use of other elements if required for local implementations.

2.0 The Dublin Core Metadata Set
The 15 elements that comprise the Dublin Core have been stable since 1996. They may be used in any order, and each element is optional and repeatable. Element names, identifiers, and definitions for Version 1.1 ¹ of the element set are as follows:

Name	Identifier ²	Definition
Title	Title	The name given to the resource
Creator	Creator	An entity primarily responsible for making the content of the resource
Subject and Keywords	Subject	The topic of the content of the resource
Description	Description	An account of the content of the resource
Publisher	Publisher	An entity responsible for making the resource available
Contributor	Contributor	An entity responsible for making a contribution to the content of the resource
Date	Date	A date associated with an event in the lifecycle of the resource
Resource Type	Type	The nature or genre of the content of the resource
Format	Format	The physical or digital manifestation of the resource
Resource Identifier	Identifier	An unambiguous reference to the resource within a given context
Source	Source	A reference to a resource from which the present resource is derived
Language	Language	A language of the intellectual content of the resource
Relation	Relation	A reference to a related resource
Coverage	Coverage	The extent or scope of the content of the resource
Rights Management	Rights	Information about rights held in or over the resource

3.0 Extending the Dublin Core
For some applications, this simple element set may not capture all the descriptive information required. Two mechanisms exist to support the capture of more complex or detailed information.

3.1 Qualified Dublin Core
To increase the specificity of the metadata, Dublin Core can be qualified. There are two forms of qualifiers. Element qualifiers refine the semantic meaning of certain elements. For example, a resource may have two significant dates, such as the original date of publication of a text and its date of publication on the Web. In order to avoid ambiguity, these two date expressions can be qualified as DC.Date.Created and DC.Date.Issued respectively.

Secondly, value qualifiers refine the value ascribed to an element within a given metadata record. These may specify a standard encoding scheme to which the value conforms or a controlled vocabulary from which the value is drawn. For example, a date value such as 1999-09-12, qualified as encoded according to ISO Standard 8601, allows this date to be understood as September 12 not December 9, 1999.

Until recently, the qualification of Dublin Core (DC) was at the design and discretion of implementors. However, it was recognized that the greater degree of non-standardized qualification, the greater the potential loss of interoperability ³ . Therefore, much of the recent work of the Dublin Core Metadata Initiative has focused on standardizing the most common approaches to qualifying elements and values. At the seventh Dublin Core Workshop (held in October 1999 in Frankfurt, Germany), a limited number of general-purpose qualifiers were proposed to become DC-sanctioned qualifiers, to be managed within the DC namespace. The Dublin Core Metadata Initiative will issue a formal set of sanctioned qualifiers by January 1, 2000.

3.2 Additional Elements
It is expected that Dublin Core will co-exist with other metadata standards and that local requirements may require additional extensions. Thus a second method to capture more information is by adding elements drawn from other standard element sets or developed locally to meet specific search and retrieval needs. For example, an implementor could extend the element set by adding a non-Dublin Core element such as Audience if the intended level of user were seen as a significant factor for finding the resource or distinguishing it from others. Another implementor might add a block of metadata drawn from another scheme (such as <indecs> for rights management data). To identify these as extensions beyond the Dublin Core, the "DC" prefix commonly preceding the element name is replaced by another appropriate prefix. At the present time, such extensions are not sanctioned nor maintained by the Dublin Core Metadata Initiative.

4.0 Encoding Syntax
The Dublin Core standard is concerned with semantics – what the elements mean. How the elements are expressed for humans or computers is a question of syntax. Syntax issues are not part of the standard.

The relationship between metadata and the resource it describes may take one of two forms: elements may be contained in a record separate from the item, or the metadata may be embedded in the resource itself. If embedded, there are two established conventions for expressing the metadata (be it Dublin Core or another type): the HTML <META> tag syntax, or the Resource Description Framework (RDF). The former has been more commonly used to date; the latter, which is an expression of the eXtensible Mark-up Language (XML), is more prescriptive in the way the data must be written, and therefore more readily supports machine-readability and the union of data across applications.

The Dublin Core standard prescribes neither the relationship of the metadata to the resource it describes nor the syntax for expressing the metadata. Nevertheless, the Dublin Core community has developed specific Dublin Core-based conventions and guidance to support standardized use of both of <META> and RDF syntaxes, and there is increasing consensus that the latter is the preferable syntax.

5.0 Strengths
The growing importance of the Dublin Core for electronic resource description may be attributable to several key characteristics:

Simplicity – It is intended to be simple and intuitive enough for content creators to describe their resources without special training.

Extensibility – To support richer description, elements can be repeated or qualified. Local extensions are not precluded.

Semantic interoperability – Each element means the same thing across disciplines.

International consensus – The scheme derives from the collective wisdom and experience of practitioners in a range of fields from over 20 countries worldwide.

Multilingualism – The element set and its supporting documentation have been translated into 24 languages to date, and this work is ongoing.

Flexibility – While the Dublin Core initiative began with a focus on electronic resources, work has been invested to ensure the scheme is sufficiently flexible to represent resources (and relationships among resources) that are both digital and exist in traditional formats.

Formal standardization – The scheme is currently an Internet Engineering Task Force (IETF) Recommendation and will be submitted to the next-level standards bodies, the American National Information Standards Organization (NISO) and the European Centre for European Standardization (CEN) this year.

Adoption within related standards – The Dublin Core has recently been adopted by the Open eBook Publication Structure 1.0, which is the industry standard to support the production of device-independent commercial electronic books. It is also used as the basis for recent proposals for administrative, collection-level, and record-keeping data sets, as well as having been used within the Gateway to Educational Materials (GEM) scheme and Visual Resources Association metadata.

Administrative infrastructure – The Dublin Core Metadata Initiative has formalized its committee structure, developed an extensive Web site and a range of open discussion lists, and implemented a formal ratification process. The Online Computer Library Centre’s (OCLC’s) Office of Research and Special Projects currently hosts the Dublin Core directorate.

Implementation experience – The Dublin Core has achieved significant acceptance within diverse sectors including archives, national governments, museums, libraries, and universities, and among subject communities including the arts and humanities, bibliography, business, education, environment, mathematics, medicine, science and technology. This uptake has resulted in an growing body of models, implementation guidance, and software tools.

Mechanism for ongoing evolution – The work of the Dublin Core Metadata Initiative is conducted in open working groups on specific issues and at the annual invitational workshops.

6.0 Challenges
To critics of the Dublin Core element set, some of its core characteristics are also its inherent weaknesses. Indeed, uncontrolled usage has compromised interoperability across applications. Semantic interoperability has been elusive because element definitions have been stretched and at times misinterpreted by implementors. It is hoped that the current effort to standardize qualifiers will result in reduction in the amount of local interpretation and customization.

More implementation experience, useful tools to create and manage Dublin Core metadata, documentation and guidance for its use, and promotion of the Dublin Core are needed. But the year 2000 holds much promise; the Dublin Core community is growing, and the place of this standard within the broader metadata standards context is solidifying.

7.0 Conclusion
The Dublin Core Metadata Initiative currently enjoys broad international support from the library, cultural and computing communities. Significant progress has recently been made to resolve organizational and implementation challenges such as widespread variance in usage. The Dublin Core metadata set has a proven track record within many large resource description projects. It is now poised to prove itself as a foundation for interoperability among such projects.

8.0 Resources

Dublin Core Metadata Initiative. – Access: http://purl.oclc.org/dc/

Dublin Core Metadata Element Set, Version 1.1: Reference Description. – Access: http://purl.oclc.org/dc/documents/rec-dces-19990702.htm

List of selected Dublin Core-based projects. – Access: http://purl.oclc.org/dc/projects/

O. Lassila, R. Swick. – Resource Description Framework (RDF) Model and Syntax Specification. – 1999 – Access: http://www.w3.org/TR/REC-rdf-syntax/

E. Miller, P. Miller, D. Brickley. – Guidance on expressing the Dublin Core within the Resource Description Framework (RDF) [Working Draft]. – 1999. – Access: http://www.ukoln.ac.uk/metadata/resources/dc/datamodel/WD-dc-rdf/

S. Cox. – Recording qualified Dublin Core metadata in HTML. [Working Draft]. – Access: http://purl.oclc.org/dc/documents/notes-cox-19990816.htm

S. Weibel. – "The State of the Dublin Core Metadata Initiative April 1999." – D-Lib Magazine. – Vol. 5, no. 4 [April 1999]. – Access: http://www.dlib.org/dlib/april99/04weibel.html

__________
Notes

1. The reference description for Version 1.1 is available at http://purl.oclc.org/dc/documents/rec-dces-19990702.htm

2. Identifiers are the non-translated machine-readable index tokens.

3. Miller, Paul. "Unifying Resource Discovery for the Humanities: An Application Based Upon the Dublin Core": http://ahds.ac.uk/public/metadata/discovery.html. (cf. Cromwell Kessler…)