©2002 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Taming Metadata in the Wild WestLiz Bishoff, Colorado Digitization ProgramCheryl Walters, Utah State UniversityChuck Thomas, Florida State UniversityElizabeth S. Meagher, University of Denver
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Who what when where—Western Trails Digital Standards• Western Trails 2001 IMLS funded grant
– Multi-state initiative to create a collection of digital objects on topic of Western Trails
– 23 participating institutions, creating 20,000 digital objects
– Each institution would host their own digital object/each would create their own metadata/each would use their own metadata standards and their own database or local system
– Each state would create a statewide database
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Who, what, when, where (con’t)• Interoperability of the 4 state databases
was through Z39.50, with a SiteSearch Web-Z interface
• Based on the CDP experience– Crosswalks from various databases– Reviewed the CDP Best Practices– Agreed to utilize Dublin Core as the common
format• Involved more than just the 4 states
– Utah Academic Library Consortia, New Mexico, Arizona, Minnesota, Kansas, Nebraska, Colorado, Wyoming
– 18 representatives from archives, museums, historical societies in these states met over a 9 month period to develop the document
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Why Western States Best Practices?
• Improve user results/satisfaction• Improve consistency across different
cultural heritage institutions• Enhance potential for creating union
catalogs from multiple databases/ILS• Provide guidance for cultural heritage
institutions on use of Dublin Core for digital resources
• Support interoperability• Support emerging standards--OAI
©2002 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Taming Metadata in the Wild West
Part 2: Writing the metadata guidelines
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Western Trail Metadata Task Forces- Descriptive Elements
- Title - Contributor- Creator - Publisher- Subject - Language- Description - Source- Date.Original - Coverage- Relation
- Technical Elements- Date.Digital - Format.Creation- Type - Identifier- Format.Use - Rights Management- Holding.Institution
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Getting It Done• Set up two electronic discussion lists• Used Colorado Digital Project’s metadata
guidelines as a base document• Created initial working draft
– base document– decisions made during WSDSG’s first meeting– input from task force members
• Distributed draft to task force for revision (additions, deletions, rewrites, etc.)
• Changes discussed via email & incorporated• Result taken back to entire Western States Digital
Standards Group for review
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Some concepts• Define all terms used• Avoid being library-centric • Do not assume cataloging or
metadata experience• Provide lots of examples• Provide links to related thesauri,
standards, etc.
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Problems, points of contention• Figuring out exactly what data each
Dublin Core element should containNot as easy as it sounds!
- Figuring out how to make guidelines flexible & comprehensive enough to fit a variety of situations, collaborative ventures & partners, for now and in the future.
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
The Coverage element• Official DC definition:
“The extent or scope of the content of the resource”
• What does that mean exactly?• How does it differ from the Subject• Does not mean date/place of publication• Our description of this element:
“Describes the spatial or temporal characteristics of the intellectual content of the resource”
– For art objects and artifacts, this could be the place where the object originated and the date or time period during which it was made.
– Currently recommended only for maps, etc. or when place or time period cannot be adequately described by Subject element.
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Source versus Relation element• Looking at California State Library’s
Metadata Standards helped– Source maps to MARC 534
• Note about original version
– Relation maps to MARC 787• Note about a related title
• The lights came on for catalogers who wanted to provide similar MARC tag equivalents; voted down as too library-centric. There are other standards beside MARC!
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Relation refinements explained• Relation element has 12 possible
refinements • Meaning of each not always obvious• Some of the differences not clear
– Relation.IsFormatOf versus Relation.HasFormatOf.
- We provided DC’s explanation of the relationship between the resource and the object described in relation field.
- Also gave concrete example of each
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Publisher element• Not straightforward when object is
digitized version of previously published item– Is it the digitizing institution?– Is it the publisher of the original version?
• Our guideline explains – “The Publisher element contains information
about the digital publisher. Publisher information from earlier stages in an object’s publishing history may be listed in ... Source and Contributor.”
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Date element• What could possibly be confusing about
“Date”?- Date originally issued, published,
made, or created?- Date digitized?- Date of an associated event?
- We created two new refinements to distinguish most important dates: - Date.Original “Creation or modification dates
for the original resource from which the digital object was derived or created.”
- Date.Digital “Date of creation or availability of the digital resource.”
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Enter initial articles in titles?• Sounds innocuous but ...
Affects sorting for display or reportsDo you want the title “The toupee worn by....” to sort by “Toupee” or “The”?
- MARC controls via indicators; other formats don’t have
- If leave out, creates possible problem if migrating data into or out of a MARC format databases.
- One person dryly commented: “there will probably be some sort of trouble
no matter what we decide.”- Our guidelines recommend omitting initial
articles.
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Making guidelines“one size fits all”• Tried to encourage users to think about
ramifications of their metadata decisions• Reminded them to think about how data
may be migrated and shared in future• Listed lots of different thesauri &
schemes to give users some choices • Listed important info that metadata
creators should include in record.– Example: the Format.Creation field
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Improving quality & detail of data• Format.Creation field guidelines
describe important technical data that users might want to include:– File size, quality (bit depth, resolution), extent
(playtime, etc.), compression, checksum value, operating system, creation hardware & software, etc.
• Format.Creation also gives links to resources with more info about terms and standards.
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
A rose by any other name...• One user community’s “autograph
album” might be another’s “libri amicorum”
• How can we accommodate many potential controlled vocabularies
• Does the public need or want to know what vocabularies are used?
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Flexible subject guidelines – Allow and provide links to many
different thesauri– Separate out different subject/genre
schemes• Example: Put all the Lib of Congress subject
headings in one field; put all the genre terms from Thesaurus for Graphic Materials in another.
– Identify thesauri used via scheme qualifier in field label, not mixed in with data in field itself which is searchable.• Example: Label is Subject.MeSH so that
“mesh” does not become a searchable term.
©2002 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Taming Metadata in the Wild West
Part 3: Applications
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Metadata Application Depends On:
• Information available about the artifact
• Expertise of the researcher
• Complexity of records
• Expertise of the cataloger
• Data entry system and display
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
MARC to Dublin Core – DCBuilder
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Original Museum Record
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Museum Record after CDP
Title Acer florissantiCreator Contributor Link http://planning.nps.gov/flfo/tax3_Detail.cfm?ID=13484004 [Access] [URI]Publisher 1. Florissant Fossil Beds National Monument 2. National
Park ServiceDescription Plant (Angiosperm, Dicotyledon) Family: AceraceaeDate Digital 2000Subject(s) Aceraceae -- Colorado
Angiosperms, Fossil -- Colorado Dicotyledons, Fossil – Colorado Florissant (Colo.)
Type 1. image [DCMI Type Vocabulary] 2. text [DCMI Type Vocabulary]Source National Museum of Natural History, Smithsonian Institution USNM-333761Languages eng [ISO 639-2]Relation MacGinitie, D.D., Fossil Plants of the Florissant Beds,
Colorado, CarnegieFormat Use 1. image/jpeg [IMT] [medium] 2.text/html [IMT] [medium]Rights National Museum of Natural History, Smithsonian
InstitutionProject Florissant Fossil Beds National Monument
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Metadata Record in ContentDM
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Metadata Record in ContentDMContinued
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Metadata Elements - Public Display
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Historical Society Metadata Record
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Direct InputTitle Annual report of the Jewish Consumptives' Relief Society at Denver, Colo.Creator Jewish Consumptives' Relief Society (U.S.)Contributor Link http://library.du.edu/About/collections/SpecialCollections/jcrs/annualreports.cfmPublisher University of Denver. Penrose LibraryDescription The <3rd- > reports published <1907- > as regular numbered issues of:
The Sanatorium, v. <1- > The 11th and 12th reports (covering 1914-15) issued in combined form as: The Sanatorium ; v. 10, nos. 3/4 (July-Sept./Oct.-Dec. 1916) Reports cover the year ending Dec. 31. Chiefly in English, with some Hebrew.
Date Original 1905-1906. [Issued] [W3C-DTF]Date Digital 2002-01-04 [Created]Subject(s) Jewish Consumptives' Relief Society (U.S.) -- Periodicals.
Tuberculosis -- Patients -- Colorado. Sanatoriums -- Colorado -- Denver.
Type image [DCMI Type vocabulary]Source 23-26 cm.Languages eng [ISO 639-2]; heb [ISO 639-2]Relation Beck Archives/Rocky Mountain Jewish History Society. Jewish
Consumptives' Relief Society Collection. Special Collections Dept., Penrose Library, University of Denver, Denver, Colo.
Format Create jpg; 300 dpi; 145 files; Epson Expression 836 XL Scanner; Adobe Photoshop version 5.5.
Format Use image/jpg [Medium] [IMT]Rights http://www.penlib.du.edu/specoll/copyri.html
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Metadata Record in ContentDM
©2002 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Taming Metadata in the Wild West
Part 4: Accommodation of levels of expertise
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Local Metadata Routes
LEGACYMETADATA
constituents
mappings&
migrations
NEW
CONTENT
latestmetadatastandard
Local MetaBaseServices
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Heritage ColoradoMetadata
Colorado Western Trails
Metadata
Z39.50 Connections
Local MetaBase
Conversion Scripts
Z39.50 Access
content w/out local database
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
COLORADOHERITAGE
FSUDIGITAL
WESTERN TRAILS
SERVICES PROVIDER
CONTENTPROVIDER CONTENT
PROVIDER
OAI-WT
OAI-WT
OAI-DC
OAI-WT OAI-DC OAI-METS
MOUNTAINWEST DL
CONTENTPROVIDER
OAI Access
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Content Provider Challenges
Implementing OAI - Intermediate Brokers May Be Necessary
Choosing Brokers & Harvesters
Maintaining Current OAI Provider Support
Awareness of Current Metadata Standards
Mapping Local Metadata to Supported Schema
Maintaining Current Transformation Procedure - Examples
Knowing Who Has Your Metadata
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Service Provider Challenges
Maintaining Current OAI Harvester Support- Continuing support for older versions
Awareness of Communities & Metadata Schema- What to collect?- Multiple views / repurposing- Added value of relationships between objects/collections- Link in a greater series of brokers?
Maintaining Multiple Data About Same Objects?- Examples
Active Role as Harvester/Service Provider- Contrast with more passive current OAI role
©2003 Colorado Digitization Program http://www.cdpheritage.org©2002 Colorado Digitization Project http://
coloradodigital.coalliance.org
Thank You!Liz BishoffColorado Digitization [email protected]
Cheryl WaltersUtah State [email protected]
Chuck Thomas Florida State University Libraries [email protected]
Elizabeth “Betty” MeagherUniversity of [email protected]