Getting a handle on alphabet soup

1
Getting a handle on alphabet soup

• Implementing diverse content standards and
  controlled vocabularies for metadata creation
• Jenn Riley
• Indiana University Digital Library Program

2
I picked a metadata format!
VRA Core
EAD
MARC
CDWA Lite
MODS
Dublin Core

• We can get started now, right?

Well, no.
3
So what else do I need to do?

• More standards to decide on
• Articulate clearly the rationale behind your
  decisions
• But more importantly, design an efficient and
  effective workflow

And this is just descriptive metadata!
4
Looking back and forward

• Factors you might not have considered when making
  the decision about what primary metadata format
  to use
• Ancillary decisions to make beyond the primary
  metadata format
• Design a workflow
• Then you can get started creating metadata

5
Looking back

• Users
• Systems
• Community

6
Meeting user needs

• Too often we dont consider this perspective well
  enough
• Access points
• Search
• Browse
• Display
• General vs. specialist focus
• Metadata structure standard chosen must support
  the discovery features our users demand

7
Description and delivery systems

• How much customization can you do?
• Does it really support the metadata format you
  want?
• Who can do any necessary customization - metadata
  staff? development staff?
• Must consider
• Metadata entry screens
• Search browse indexes
• Labels for end-user interface (hint element
  names are often not the best labels to use)

8
Participating in the community

• Are there established models to follow?
• Dont reinvent the wheel if you dont have to
• Who else is working in this area?
• Or is leadership needed in this area?
• Will your decisions result in an overall rise in
  the knowledge level in the community?
• Can you share your decisions and rationale?
• Even better, can you share schemas, code, etc.?

9
Looking forward

• Content standards
• Controlled vocabularies
• Syntax encoding schemes
• Required vs. optional elements
• Sources of information
• Matching staff with the metadata creation task
• What discovery functions the metadata will
  support
• Metadata sharing

10
Content standards

• How will you fill in the fields in the metadata
  format youve chosen?
• Can you adopt cataloging rules from a widely used
  standard? AACR2? DACS? CCO?
• What about smaller standards like the DCRM series
  or Betz?
• Where will you need to deviate from the content
  standard?
• What data elements do you need that the content
  standard doesnt cover?

11
Controlled vocabularies

• What fields will you place under vocabulary
  control?
• What controlled vocabularies (large or small)
  will you use for each of these fields?
• Will you perform new authority work?
• Will you contribute this back to the community?
• Will you maintain a local vocabulary? (either a
  full authority file or a simple pick list)
• What format will you store authority data in?

12
Syntax encoding schemes

• What fields need syntax control but not a fully
  enumerated list?
• What discovery or description function does the
  data in this field need to support?
• Can the metadata entry tool enforce your chosen
  syntax?
• Before the cursor leaves the field?

13
Required vs. optional

• Required if applicable is overused
• Be realistic when creating your requirements
• Can your system
• Display short and long record forms to
  metadata creators?
• Give a visual indication of which fields are
  required?
• Enforce or give feedback on conformance when
  saving/leaving a record?

14
Sources of information

• How hard should metadata creation staff try to
  fill in a field value?
• Just information on the item?
• Other documentation provided with the collection?
• Prescribed list of reference sources?
• Original research?
• This is a difficult balance
• Especially for unique materials, your institution
  knows more about this item than almost anyone
  else
• But how much time can we really spend on a single
  item?

15
Matching staff with the task

• Dont assume the entire descriptive metadata
  record has to be created in one pass
• Use individuals to the best of their ability
• Lower-level staff for transcription
• Specialist knowledge of the materials only when
  necessary
• Automate as much as you can
• A system can help
• Default values
• Duplicate record function
• Different views of a record for different
  metadata creation staff

16
Planning for discovery functions

• Provide guidance for each field having already
  determined
• What search/browse indexes it will be used in
• When it will be displayed to users
• Make sure metadata creators know how the metadata
  will be used
• Take a long-term view when making these decisions

17
Metadata sharing

• What will this record look like outside of your
  local environment?
• Can it automatically be transformed into a
  shareable record?
• Appropriate standards
• Complete context
• Free of local jargon/codes and information only
  useful locally?

18
Pulling it all together

• Write a cataloging manual for this format and
  this metadata creation interface
• Give metadata creation staff enough context to
  understand why
• Gives you a place to document your decisions
• Youll reuse this work later - it does become
  easier as you gain experience!

19
For more information

• jenlrile_at_indiana.edu
• These presentation slides lthttp//www.dlib.indian
  a.edu/jenlrile/presentations/princeton2008/prince
  ton2008.pptgt