Dublin Core Metadata Style Guide

 

This document is intended for use as a general guideline for creating metadata records for digital resources at the University of South Carolina Libraries. These policies should be consulted when a new project is begun. Modifications can be made on a project-by-project basis, as approved by the project leader and/or Head of Digital Activities. This document may be updated to include these modifications.

This style guide is based on the Dublin Core element set. The Western States Dublin Core Metadata Best Practices, version 2.0 has been an invaluable resource in shaping the USC Libraries’ policies for creating metadata records for digital resources.

In this document:

[  FILE NAMING SCHEME  ]  [  DUBLIN CORE ELEMENTS FOR DESCRIPTIVE METADATA  ] [  ADMINISTRATIVE METADATA  ]  [  PRESERVATION METADATA  ]  [  REFERENCES  ] 



FILE NAMING SCHEME

The first step when starting a new digital project is determining the file naming scheme to be used for that project. All schemes should include only alphanumeric lowercase characters and will not include spaces, except that the “_” underscore may be used. There can only be 8 characters before the 3 character file type.

The Head of Digital Activities and/or the collection administrator will determine the scheme, according to the following model:

Library Dept. code (1 character) +
Collection # (3 characters, numeric order of collections scanned for department) +
Item # (3-4 characters, may or may not include information about the content)
3-letter file type (.jpg, .tif, etc.)
For example: r0010001.jpg


LIBRARY DEPT CODES
Rare Books and Special Collections    r
South Caroliniana Library s
Maps Library m
Music Library u
Newsfilm Library n

The file naming scheme will be described in the administrative metadata for the collection.

DUBLIN CORE ELEMENTS FOR DESCRIPTIVE METADATA

Title  |  Title.Alternative  |  Creator  |  Subject  |  Description  |  
Publisher  |  Contributor  |  Date.Original  |  Date.Digital  |  Type  |  Format  |  
Digitization Specifications  |  Resource Identifier  |  Source  |  Language  |  
Relation  |  Coverage  |  Rights Management  |  Contributing Institution  |  Web Site

    Title

  • Each record will only have one title
  • Follow AACR2 rules for capitalization, punctuation, etc.
  • Remove all leading articles. For example: The, A, An
  • If no title, title supplied by cataloger, but do not put title in brackets. Project leader determines conventions for supplying titles.
    Title.Alternative
  • Enter each Title.Alternative in one field, separated by a semicolon and one space
  • Alternately, enter Title.Alternatives in separate field, but with unique label names (e.g. Title.Spine, Title.Subtitle, etc.)
  • Follow AACR2 rules for capitalization, punctuation, etc.
  • Remove all leading articles. For example: The, A, An
  • Use for subtitles, titles with ampersands spelled out, etc.
    Creator
  • Enter each Creator in one field, separated by a semicolon and one space
  • Follow AACR2 rules for capitalization, punctuation, etc.
  • Use LC Authority File for form of corporate and personal names
  • Enter personal names in inverted form: last name, first name. Enter corporate names in full direct form.
  • If no LC established heading exists, project leader makes decision on name
  • Can include multiple names or corporate bodies, but also use Contributor field for those who played a secondary role in the intellectual content of the item
  • Include role of creator, when needed, in parentheses after name. Will make list of roles and decide whether to use codes on a project by project basis (e.g. photographer, author, printer, etc.)
    Subject
  • Enter each Subject in one field, separated by a semicolon and one space
  • Use controlled vocabularies- LCSH & AAT, or local when necessary
  • Use the following labels for online display: Subject.LCSH, Subject.AAT, Subject.Local
  • Can add terms when relevant – project leader approves
  • Can use USCAN as guide for inputting LC subject headings
  • Local subject headings are determined by project leader
    Description
  • Enter each Description in one field, separated by a semicolon and one space
  • Use for free-text account of the intellectual content of the original item, and physical description of original item. Can be quoted from item itself, or composed by metadata creator
  • Is searchable, so when inputting, think of audience and terminology they would likely use
  • Use both Description.Abstract and Description.TableofContents, where applicable (Abstract is composed by metadata creator, but TableofContents is found on item)
    Publisher
  • Each record will only have one publisher: University of South Carolina
    Contributor
  • Enter each Contributor in one field, separated by a semicolon and one space
  • Follow AACR2 for capitalization, punctuation, etc.
  • Use LC Authority File for form of corporate and personal names
  • Use this field for entities that had a secondary role in the intellectual content of the original item, and so are not listed in Creator field
  • Can include role of contributor in parentheses after name; use on a project by project basis. List of roles will be provided by project leader
    Date.Original
  • Each record will only have one Date.Original
  • Follow ISO 8601 standard of YYYY-MM-DD, YYYY-MM or YYYY. Only use what part of the date is known. For example, if only the month and year is known, use YYYY-MM
  • Use date(s) associated with the creation or availability of the original item, not of the digital version
  • Always make educated guess, so that there are very few missing dates
  • If missing date, use: n.d., or if approximate date known, use: circa YYYY
  • For range of dates, put space-dash-space in between dates. For example, 1940 - 1945
  • Make refinements if necessary, on a project by project basis
    Date.Digital
  • Each record will only have one Date.Digital
  • Use for the date the original item was scanned, or the date the digital item was imported onto server
  • Follow ISO 8601 standard of YYYY-MM-DD, YYYY-MM or YYYY
    Type
  • Each record will only have one Type
  • Follow DCMI Type vocabulary. For example: Still Image, Moving Image, Sound, Text
    Format
  • Each record will only have one Format
  • Consolidate IMT (Internet Media Type) and subtype into one text string
  • Add file size and content extent when applicable
  • For example:
    • image/jpg 2 MB (for a digital photograph), or
    • video/mpg 10 GB ; 34 min., 20 sec. (for digital video)
  • Do not follow Western States guidelines about bytes. Use MB, KB, GB, etc., as appropriate
    Digitization Specifications
  • Each record will only have one Digitization Specifications
  • Use on project by project basis
  • Include: bit depth, resolution, scanner hardware and software used and archival master file type (ie: tiff)
    Resource Identifier
  • Each record will only have one Resource Identifier
  • Use for file name
    Source
  • Enter each Source in one field
  • Use for free-text account of the source of the original item
  • Include publisher of original item here, plus ISBN or any other original necessary information
  • Maintain detail from MARC catalog record when possible, especially for manuscripts
  • Can also include URL of finding aid or MARC record in USC’s online catalog
    Language
  • Enter all Languages in one field
  • Spell out the language. For example: English
  • If more than one language, separate by a semicolon and space. For example: English; Spanish
  • Most prominent language listed first
    Relation
  • Enter each Relation in a separate field
  • Use refinements to specify relation, when necessary
  • Always include title of Digital Collection in a Relation.IsPartOf field
    Coverage
  • Enter each Coverage.Temporal or Coverage.Spatial in one field, separated by a semicolon and one space
  • Use Coverage.Temporal for time period covered by the intellectual content of the original item (not the publication date), but use only for date and time-based events when the time element is critical for identification (For example, photograph), but there is no appropriate LCSH or AAT term
  • For Coverage.Temporal, use ISO 8601 date(s) or free-text term(s); if more than one time element is appropriate, put all entries in one text string, separated by a semicolon and space
  • Use Coverage.Spatial for geographical location(s) covered by the intellectual content of the original item (not the place of publication), but use only when there is no appropriate LCSH or AAT term
  • Use Coverage.Spatial for latitude/longitude coordinates and scale, when given, for cartographic material
  • For Coverage.Spatial, use TGN (Thesaurus of Geographic Names)’s Display term for place (or Preferred term if there is no Display term); if more than one place name is appropriate, put all names in one text string, separated by a semicolon and space
    Rights Management
  • Each record will only have one Rights Management
  • Every library or department uses own unique wording
    Contributing Institution
  • Each record will only have one Contributing Institution: University of South Carolina. [Library Department]
  • Use LC Authority File for library department name. If there is no established heading, consult project leader
    Web Site (Local field used for every record)
  • This field will contain the url for the home page of the digital collection

ADMINISTRATIVE METADATA

The following information will be maintained in a database separate from the descriptive metadata describing the collection. It will not be accessible to the public, and will be used for administrative purposes.

Project Identifier
Collection Name
Collection ID
Collection Administrator
Library Department
Description of collection
Date began project
Date finished project
Total size for storage
Number of items
Metadata persons
Scanning person
Outside faculty liaison
Proposal submitted
Proposal document
Scanner and software
Image specs (resolution, bit depth, average file size, color)
Special Texts completed
CONTENTdm main page created
DateImagesUploaded
Storage site of tiffs
Storage of jpegs
Storage of other
Archive Justification
Archive Review

Back to Top 

PRESERVATION METADATA

The following information will be maintained in the TIFF header separate from the descriptive metadata describing the images. It will not be accessible to the public, and will be used for preservation purposes.

Date Scanned Scanner Technician Title Collection Scanner and Scanning Software


Back to Top

 

REFERENCES

Dublin Core Metadata Initiative
http://dublincore.org/

Western States Dublin Core Metadata Best Practices, version 2.0
http://www.cdpheritage.org/resource/metadata/wsdcmbp/

Library of Congress Authorities
http://authorities.loc.gov/

Thesaurus of Geographic Names
http://www.getty.edu/research/conducting_research/vocabularies/tgn/

Internet Media Types (aka MIME)
http://www.graphcomp.com/info/specs/mime.html

DCMI Types
http://dublincore.org/documents/dcmi-type-vocabulary/

Art and Architecture Thesaurus
http://www.getty.edu/research/conducting_research/vocabularies/aat/

Feedback

 

Columbia Departments Campus Libraries
Columbia Libraries and Collections