(First crack at a) Presentation Outline


Purpose of Presentation:

(~1 hr, including 15 min for questions / discussion)
  • No spin, just information (seeking feedback from others, not making recommendations)
  • Opening a dialog about federated search here at RPI
    • Give people an idea of the landscape of federated search and related products - build awareness
    • What products are out there, what can they do, and how are libraries using them?
  • Engage people
  • Give people a jumping-off point to do their own investigations - provide lots of links, etc. (wiki)
  • After presentation, solicit feedback on whether we should look into federated search more.
    • Encourage people to make comments and ask questions throughout.
      • (work out timing after have presentation guts more put together)
  • Make a handout

Presentation content:

  • What's the motivation for federated search? (MB)
    • Users increasingly look for a single search box. Federated search tries to give this to them, with varying amounts of success
  • Start with definitions / clarification. (MB)
    • what's the difference between discovery layer / metasearch / federated search? - fuzzy definitions, no agreement on terminology.
      • even within one vendor, it can be hard to tell what product does what. what are you buying,exactly.
    • give scope/range of what they can and can't search.
    • overall, how would this help the user?
      • incorporate matt's usability studies?
  • What products are out there? (KD)
    • (Wouldn't want to spend a lot of time listing products and their features - hard to understand / remember. Try to show different products as examples when discussing general concepts/features instead?)
    • List of products we've looked into may not be exhaustive, but we did get the big ones.
      • List these for people on a slide / handout and refer people to wiki for more info
    • There is a lot of commonality of features between different products
      • They all make similar claims, but differences lie in how they do them and how well?
    • Open source options
      • uncertain future for support of two open source federated search tools we looked at. Concerns about committing to a product during a time of transistion.
        • when something is no longer supported / being actively developed:
          • you don't get new features
          • fewer other libraries using it who you can troubleshoot and share ideas with
  • What do they do / features? (Lots of examples here) (KD)
    • what can be searched? this varies between products
      • not all products let you combine catalog results w/ article database results
    • de-duping
    • sorting of results (relevance ranking, date, other criteria?)
    • relevance ranking
    • narrowing of results by facets, or by source database.
    • ease of identifying source database and running your search there ("More from Academic Search Premier")
    • some return results faster than others
    • Integration with subject pages
      • kind of like mylibrary/libguides? (example: metalib server)
    • Compatibility with other products
      • what they can search
        • example: Explorit (from Deep Web / Stanford) says it can search InMagic databases, which is what Archives' campus publications / images database is in.
        • example: several can search OAI-harvestable repositories (like Digitool)
      • incorporation into discovery layers
      • integration with SFX to link to full-text
    • Identification of full-text results
    • Does it search a locally-created index or send the query out to the databases in realtime and return results from there?
What might we want to combine in one search at RPI? (TAG)
  • Databases.
  • Library catalog
  • e-journals (list of titles)
  • Archives
    • campus pub / image database (inmagic)
    • finding aids on the web (HTML, PDF, (EAD?))
    • web resources (HTML)
  • Digitool
  • Architecture
    • Saskia
    • others?
  • web pages / locally-created content?

  • End with a series of steps we'd need to take WERE WE to investigate federated search further? (TAG)
      • Further investigation / feedback from library staff, users.
      • Choosing a product/implementing (IF implementation were to take place after further investigation)
        • would need to make sure we have the resources we need(people, $, technical)
        • decide what information resources would we want a federated search to cover?
        • Features that are important to us.
        • how would we want to present it to users?
  • Solicit conversation / questions at the end - 15 min?
    • be ready with some questions to prompt conversation if there aren't many from the audience.
    • How can we encourage this?

After presentation:

    • Gather feedback
      • Use this wiki for this? (A page? Discussion?) Ask Fran how folks are using the ref wiki & whether she thinks a page or discussion would work better.
    • Decision will need to be made on whether to move forward w/ investigating the possibility of using federated search at RPI. -
      • is it even worth investigating if we don't know whether we'll have the resources (human, financial, technical?) to support it?

Things we need to clarify before presentation:

    • What was William's Google Scholar idea?
    • Forgot to tell Tammy & Matt - I'm getting Adobe Connect training to maybe involve Hartford
    • CLARIFY pre-presentation:
      • what are best modes for gathering feedback?
      • how does the decision get made to further explore this or not?
        • Ask Bob about this. Is it him, tech committee, librarians, management team? some combination of those?
        • Another Bob question: Funding: is the Fixman fund for things like this?