lauren's library blog

Andrew Nagy, Serials Solutions
Scott Garrison, Western Michigan University

• OPAC silo…
  • Hasn’t kept up with Web, users’ expectations
  • Limited customization
  • Antiquated, rigid search technologies
  • Designed for known-item searching
  • Libraries have set expectations, learned to compensate accordingly
• Ejournal and database silos
  • More every year in multiple packages
  • More alternatives, more confusion
  • Multiple A-Z lists to maintain, use
  • Interfaces change regularly
  • Query syntax varied, requires instructions???
  • “The version of ____ I teach is ______”
• Showed screenshot of full text options page, so many choices to click on…
• Cross-silo federated search
  • Allows some general, discipline searching
  • Mixed, incomplete results
  • As slow as the slower silos
  • If local, very network-inefficient
  • Many different metadata schemas, less sophisticated searching
• Changing Marketplace
  • Vendor acquisitions, consolidation, catch-up
  • Open source options are emerging
  • Some products are still years away
  • All of the above leads to great FUD
• Discovery as a way to gain sight or knowledge of
• Discovery layer
  • Searching for the 21st century
  • Built on 21st century technology
  • Highly configurable interfaces
  • Puts our metadata to better use
  • Works for OPAC and other silos but relies on federated search, through evolving
• Next Generation Catalog: what does it do?
  • Provide simple, easy access to the library’s local collections
  • Supplements “classic” OPAC
  • Refines searches with “facets”
  • etc
• Opensource: VuFind, Blacklight, eXtensible Catalog
• Commercial: AquaBrowser, WorldCat Local Primo, Encore, Endeca
• VuFind
  • Mellon award for Technology Collaboration in 2008
  • ILS-agnostic, runs alongside OPAC
  • Libraries of all sizes
  • Feature rich
• Today’s students aren’t afraid of iteration. They iterate again and again (like a game) until they get their answer.
• VuFind implementation at WMU
  • Themes from usability testing: fewer failed searches, user less likely to give up searching, users curious about things like tagging.
  • VuFind lets library define relevancy and set of indexes
  • Some users don’t get facets or how to use them
  • Recall -> huge adjustment from librarians
  • Takes only good records; prompted rethinking of approaching work from r
  • Includes a “spot an error” link in the catalog results to point out when shouldn’t get the result from that page (maybe a bit different from our general feedback link)
• One compelling starting place
  • Presearch in Amazon, Google, del.icio.us
  • Then they use the library catalog to find things
• Services we could provide to enhance this
  • incorporate everything they need to search in a simple interface
  • Local index of collections: MARC, OAI, etc
  • Customizable
  • Mashups
  • Tuned relevancy ranking
  • Facets
  • Citation management tools
  • Links to value-adds like ILL, recommenders
• Discovery can go further
  • Why only local collections?
  • What about article content?
  • What if users want to discover items outside their discipline-specific databases?
  • Can’t we do better than federated search?
• Web-Scale Discovery aka “Unified Discovery Service”
  • Unifies local and subscription content
  • Web-scale repository
  • Highly tuned relevancy
  • Pluggable API for “shopping mall” access; access it from a number of places on the web
  • Software as a service rather than a product
• Summons from Serials Solutions
• In use at WMU
  • Even bigger adjustment for library staff
  • Has reminded the library of record problems
  • Shows known OpenURL target problems
  • How to present it along with VuFind?
• Keep NGC for collection and summons for everything else.
• At WMU: when search VuFind with no results, offered Summons as an alternative.

Related posts:

1. NCSU libraries rock!
2. visual catalogs
3. keynote Session 2
4. Google’s Phone Solution
5. PennTags!