Is Dublin Core Dying? 

Kayla Willey – Brigham Young University

Cheryl Walters – Utah State University 

Utah Library Association Annual Conference

St. George, Utah

May 17, 2006

 
 
 
 

What is Dublin Core? 

• Title
• Creator
• Subject
• Description
• Publisher
• Contributor
• Date
• Type
• Format
• Identifier
• Source
• Language
• Relation
• Coverage
• Rights
• Dublin Core
• http://dublincore.org/documents/dces/

 
 

• Title (mandatory)
• Creator (mandatory if available)
• Subject (mandatory)
• Description (mandatory)
• Date Digital (mandatory)
• Date Original (mandatory if applicable)
• Format (mandatory)
• Digitization Specifications (mandatory)
• Resource Identifier (mandatory)
• Rights Management (mandatory)
• Publisher
• Contributor
• Type
• Source
• Language
• Relation
• Coverage
• Contributing Institution
• CDP Metadata Working Group Dublin Core Metadata Best Practices
• http://www.cdpheritage.org/cdp/documents/CDPDCMBP.pdf

 
 
 
 

UALC – Where have we been? 

• Adopted Dublin Core Metadata Best Practices
• Just completed an extensive document with examples of how to apply Best Practices by format
• CONTENTdm based on Dublin Core
• Aggregating server search fields (mapped to)

 Creator Date

Title Type

Subject Full text

Description Identifier

 
 
 
 

Dublin Core: An Obituary 
Jeffrey Beall, Library Hi Tech News, Nov 8, 2004, pp. 40-41 

• Dublin Core metadata standard rapidly become obsolete, never caught on, outlived usefulness
• Google essentially put it out of business
• Profit motive on the part of OCLC created by managers rather than practitioners
• Tried to be all things to all metadata users and lacked sufficient specificity to function as an operative descriptive schema

 
 
 
 

Weaknesses of Dublin Core 

• Simple system

“Its fundamental flaw is that it is designed as a lowest common denominator system.  Thus, converting from anything else (especially MARC) into DC results in a loss of specificity (and thus data) and converting from DC into anything else (especially MARC) results in woefully substandard data.”  --J. Hahn

• Scores low on interoperability (hard to crosswalk)

Examples: Initial articles in titles, single subject fields to multiple 6XX subject fields in MARC, No subfields (Young, Brigham, 1801-1877 – Family – Portraits)

• Lack of standardization of entering data

No authorized forms of names, LCSH, no required fields in GENERIC form 

 
 
 
 

Why are we still using Dublin Core? 

• Simplicity

Cost, training requirements, personnel, diverse institutional base of participants, digital production numbers expected

• Common denominator base for federated search

    Many disciplines, interest groups, and professions now developing standards

• Digital Library Tools haven’t supported other standards until recently

 
 
 
 

What’s the Answer? 

• MODS (Metadata Object Description Schema)

    Jeffrey Beal’s answer

    Strength: Intended to be able to carry selected data from existing MARC21 records and uses language-based tags rather than numeric ones

 http://www.loc.gov/standards/mods/

• METS (Metadata Encoding and Transmission Standard)

    Strength: Allows for multiple standards to exist within a single metadata record. Employs XML schema language.

    “A new infrastructure would therefore be needed to support the hierarchical representation of data required in a METS environment”  -- Final Report, CDP Metadata Standards Working Group Task Force on METS (May 2006)

 http://www.loc.gov/standards/mets/

• For the present, current tools in use point to DC (Dublin Core), but as tools mature, more is possible

    Encourage DiMeMa to offer other standards? Move? Require more machine produced metadata?