Teaching Old Dogs New Tricks Descriptive metadata in libraries and archives

1
Teaching Old Dogs New Tricks Descriptive metadata
in libraries and archives

• EAD, MARCXML, MODS, OAI
• Merrilee Proffitt, RLG

2
Why is descriptive metadata (still) important?

• Important for indexing, discovery and
  identification of resources
• Lots of old standards with new twists
• New standards with old twists

3
Whats new in the old world?

• The World Wide Web
• XML (eXtensible Markup Language), particularly
  XML Schema Language
• Playing nicely with others
• Continued (or increasing) diversity of materials
• Trends in scholarship and teaching towards using
  primary sources

4
Differences, similarities

• EAD, MARCXML, MODS are
• Data structure standards
• Expressed in XML (DTD, XML Schema Language)
• OAI is
• Data interchange protocol
• Relies on XML Schema Language

5
Encoded Archival Description (EAD) a new dog on
the block
If you feel irritable take Cas-car-ria William
H. Helfand Collection of Pharmaceutical Trade
Cards New York Academy of Medicine
6
EAD Encoded Archival Description

• Expressed as an SGML/XML DTD
• Society of American Archivists
• Supports archival descriptive practices and
  standards
• Supports discovery, exchange and use of data

7
Before EAD

• MARC AMC
• Limitations of MARC AMC
• Need for a machine readable format
• Need for platform and software independence

8
Brief history

• 1993 FindAid Project
• 1995 Alpha
• 1996 Beta
• 1998 Version 1.0
• 2002 Version 2002

9
What EAD is not

• Content or data value standard
• ISAD(G)
• APPM
• RAD
• CUSTARD?
• Archival management system

10
Continued Relationship with MARC

• Desirability of having collection-level
  description in the catalog
• Use of encodinganalog attributes
• Produce MARC records from finding aid
• Produce at least a partial finding aid from an
  existing MARC record

11
Whats new with EAD2002

• Changes suggested after years of real life
  useage
• Internationalization
• Numerous changes made to bring EAD into line with
  ISAD(G)
• Elimination of Anglo-centric semi-closed
  attribute lists
• Changes to accommodate non traditional archival
  materials

12
MARCXML an old dog gets a facelift?
Our Police Dog on Nov. 4/24 Election Day Jesse
Brown Cook Scrapbooks The Bancroft Library UC
Berkeley
13
MARCXML MARC21 in XML

• Initiative of Network Development and MARC
  Standards Office at LC
• Motivated by LCs move forward into XML
• Expressed in XML Schema Language
• Builds on earlier effort to render MARC into a
  set of SGML DTDs (1995-1998)

14
MARCXML advantages

• Simple and flexible framework
• Seamless MARC 2709 to MARCXML roundtrip
• Ease of presentation via stylesheets
• Ease of transformation via XSLT
• Can validate MARC records against the schema
• XML supports Unicode

15
From the MARCXML web site www.loc.gov/standards/m
arcxml
16
MARCXML example
17
MODS son of the old dog
Two black and white puppies, one facing the
camera and one walking away Chicago Daily News
negatives collection Chicago Historical Society
18
MODS Metadata Object Description Schema

• Initiative of Network Development and MARC
  Standards Office at LC
• Motivated by LCs move forward into XML
• Expressed in XML Schema Language
• OAI, METS, ZING
• MARC-like
• Allows for more richness than Dublin Core without
  enforcing MARC

19
MODS top level elements

• Title Info (mandatory)
• Name
• Type of resource
• Genre
• Origin Info
• Language
• Physical description
• Abstract
• Table of contents
• Target audience

• Note
• Subject
• Classification
• Related item
• Identifier
• Location
• Access conditions
• Extension
• Record Info

20
MODS features

• Does not require or assume a particular
  descriptive standard
• Intuitive design
• Element descriptions are repurposed throughout
  the schema
• Language-based element names
• Recursive hierarchy allows for description of
  complex digital objects

21
lttitleInfogt lttitlegtWilliam P. Gottlieb
Collectionlt/titlegt lt/titleInfogt ltrelatedItem
type"constituent"gt lttitleInfogt
lttitlegtPortrait of Charlie Parker and Tommy
Potter, Three Deuces, New York, N.Y., ca. Oct.
1947lt/titlegt lt/titleInfogt ltnamegt
ltnamePartgtGottlieb, William P.lt/namePartgt
ltnamePart type"date"gt1917-lt/namePartgt lt/namegt
ltidentifier type"local"gtLC-GLB23-0542lt/identif
iergt lt/relatedItemgt
22
MODS and MARC

• Does not include the full element set
• Does not cover all content rules
• Round-trip conversion not possible a one-way
  ticket only

23
MODS and Dublin Core

• MODS is richer than Dublin Core
• More natural fit in a library environment
• Unqualified Dublin Core still a excellent
  transfer syntax between diverse descriptive
  communities

24
More MODS

• Version 1.2 June 2002 January 2003
• Version 2.0 released in February 2003
• Version 3.0 now available in Draft
• Part of the MARC tool kit
• User guidelines issued earlier this year, very
  MARC based

25
How applicable for other communities?

• Based on the needs of the library community, but
  could well be useful elsewhere
• Good at describing granular information
• More generalize user guidelines will help make
  this more palatable for some

26
Open Archives Initiative new tricks for dogs
Fun-Time Drawing Diana Korzenik Collection of Art
Education Ephemera Huntington Library, Art
Collections, and Botanical Gardens.
27
What is OAI

• Open Archives Initiative
• Governing body, organization
• Open Archives Initiative Protocol for Metadata
  Harvesting (OAI-PMH)
• Mechanism for achieving the goals of the
  organization

28
OAI Mission

• develops and promotes interoperability
  standards to facilitate the efficient
  dissemination of content.the technological
  framework and standards are independent of
  the type of content offered and the economic
  mechanisms surrounding that content.

29
Some terminology

• Open
• In terms of architecture, not free
• Archives
• Not meant to be taken literally

30
OAI History

• Began in 1999 with several communities E-Prints
  servers, Digital Archives, Networked Consortia
• Led to the OAI Protocol for Metadata Harvesting
  (OAI-PMH, now at version 2.0)
• Support from a the Digital Library Federation,
  the Coalition for Networked Information, and the
  National Science Foundation

31
OAI Key Concepts

• Interested in promoting sharing of metadata among
  and between communities
• Electronic or analog resources
• Just descriptive data, not content
• Reliance on Unqualified Dublin Core and XML

32
Role of Providers

• Data Providers
• Expose metadata for harvesting
• Service Providers
• Harvest metadata and provide value-added services

33
Why Should I Care About the OAI?

• Utilized by many digital repositories
• Supported by both Dspace and FEDORA
• Has been embraced by (U.S.) federal and
  international agencies that apportion funding and
  grants (examples IMLS in the US, JISC in the UK)
• Provides a model for how information can be
  distributed to search services

34
Example OAI-PMH Request

• http//arXiv.org/oai2?verb
• GetRecordidentifieroai
• arXivcs/0112017metadataPrefix
• oai_dc

35
Example OAI-PMH Response

• lt?xml version"1.0" encoding"UTF-8"?gt ltOAI-PMH
  xmlns"http//www.openarchives.org/OAI/2.0/"
  xmlnsxsi"http//www.w3.org/2001/XMLSchema-instan
  ce" xsischemaLocation"http//www.openarchives.or
  g/OAI/2.0/http//www.openarchives ETC. ETC.
  ETC.

36
What does OAI not do?

• Rights management
• Explicit support for metadata formats other than
  Unqualified Dublin Core
• Provide access to content

37
Recap

• EAD implemented in XML, optimized for
  (international) archival community
• MARCXML implemented in XML Schema Language,
  very flexible structure, optimized for working
  with MARC data (as it is and as it will be)
• MODS implemented in XML Schema Language, very
  flexible structure, optimized for simple to
  detailed resource description in a library
  setting
• OAI simple resource discovery protocol,
  designed to work with XML Schemas

38
Questions?

Book titled ''Hope Commission Report'' is thrown
to a pack of dogs Digitised images from the
Pictorial Collection National Library of Australia