Next Generation Search Implications for Cataloging and Metadata Creation

1
Next Generation Search Implications for
Cataloging and Metadata Creation
2
Next Generation Search is already here!!!
3
University of Minnesotas Primo
4
Oklahoma State Universitys Aquabrowser
5
University of Washington WorldCat Local
6
Record within WorldCat Local
7
Record within UIUC Online Catalog
8
Amazons bibliographic records
9
Amazons rich metadata
10
Online catalog to facilitate successful search

• Search to help a researcher find what she or he
  wants
• Need to provide 1) broad net to find all that
  is out there, 2) narrow down to relevant
  resources, and 3) a way to do known item
  searching
• Intuitive navigation between results
• Need to allow the user to figure out the easiest
  way to get this content, whether full-text book,
  article, print resource, archive, dataset, image,
  etc.

11
Next Gen search can allow

• Integration of formats (and the accompanying
  metadata)
• Better inclusion of table of contents (as part of
  MARC record or as separate pile to search)
• Relevancy ranking of terms, names and phrases
  within the index of book or the actual content of
  the full-text book
• Interconnectivity of resources moving from
  citation to citation (as more and more resources
  are in digital form)

12
More and more metadata at our fingertips

• Digitization projects Google, Microsoft, Open
  Content Alliance, Amazon, Million Book Project,
  EU digitization project, etc.
• ONIX feeds from publishers and vendors
• Using METS and ATOM to create file structure to
  pull these desperate data elements together
• Using OAI-PMH or OAI-ORE to harvest data from
  various sources
• OCLCs work to provide comprehensive
  uber-metadata records for all resources,
  combining data from libraries and publishers to
  create

13
Googles More about this book metadata record
14
Facets

• Material Type
• Author/Creator
• Date
• Subject
• Library Location
• Language
• Call Number

• Availability
• Series
• Serial title changes
• Translations
• Works cited
• Keyword clouds
• Dewey has great potential as a facet

15
Facets can utilize MARC to its fullest
capabilities

• Fixed fields
• Series statements (440, 490, and 830 fields)
• Uniform titles (240) linking translations
• Language (041)
• Identifiers (022, 035, 776)
• Publisher
• Author fields (1xx, 7xx
• Authority records play an important in all this
• Accuracy is so critical to this process

16
Authority Control errors front and center
17
MARC cant handle all data

• Images
• Value-added information such as biographies,
  reviews, automation of inclusion of TOCs
• Indices
• Bibliographies
• Cross references within sources
• Correction of past practices (not including all
  the authors, abbreviations, publication
  information, etc.)

18
Catalog as Dataset

• UIUC Catalog is probably about 4 gigabytes of
  data
• Could fit on one of these external hard drives or
  thumb drive
• Robust editing tools out there for XML-based data

19
Automating Metadata Creation and Maintenance

• This is where we need to get as soon as we can
• There will always be a human element to this.
• But, we need to be able to pull in/harvest data
  from where we can.
• Need to have the ability to work with MARC and
  XML
• Need to be able to pull out a hunk of data and
  change certain elements, en masse.
• MARC can still be a critical element in this work

20
MARC (redux)

• LCSH, LCC, Dewey, Authority Control, AACR2 or RDA
  all still have an important role today but so
  does Dublin Core, MARC XML, MODS, other
  controlled vocabularies, semantic web, metasearch
  and federated search
• Keeping our surrogate metadata records as
  accurate, current and complete as possible should
  be our focus as we continue forward
• MARC still has a role to play but it needs to
  become more flexible and fluid to incorporate new
  and emerging technologies.
• These next generation search tools, such as
  Aquabrowser, Primo, WorldCat Local, VU Find,
  etc., are also critical in moving us forward to
  reclaim our central role in facilitating research