Your SlideShare is downloading. ×
0
Harvesting From Many Silos at Web-scale Makes E-content Truly  Discoverable
Harvesting From Many Silos at Web-scale Makes E-content Truly  Discoverable
Harvesting From Many Silos at Web-scale Makes E-content Truly  Discoverable
Harvesting From Many Silos at Web-scale Makes E-content Truly  Discoverable
Harvesting From Many Silos at Web-scale Makes E-content Truly  Discoverable
Harvesting From Many Silos at Web-scale Makes E-content Truly  Discoverable
Harvesting From Many Silos at Web-scale Makes E-content Truly  Discoverable
Harvesting From Many Silos at Web-scale Makes E-content Truly  Discoverable
Harvesting From Many Silos at Web-scale Makes E-content Truly  Discoverable
Harvesting From Many Silos at Web-scale Makes E-content Truly  Discoverable
Harvesting From Many Silos at Web-scale Makes E-content Truly  Discoverable
Harvesting From Many Silos at Web-scale Makes E-content Truly  Discoverable
Harvesting From Many Silos at Web-scale Makes E-content Truly  Discoverable
Harvesting From Many Silos at Web-scale Makes E-content Truly  Discoverable
Harvesting From Many Silos at Web-scale Makes E-content Truly  Discoverable
Harvesting From Many Silos at Web-scale Makes E-content Truly  Discoverable
Harvesting From Many Silos at Web-scale Makes E-content Truly  Discoverable
Harvesting From Many Silos at Web-scale Makes E-content Truly  Discoverable
Harvesting From Many Silos at Web-scale Makes E-content Truly  Discoverable
Harvesting From Many Silos at Web-scale Makes E-content Truly  Discoverable
Harvesting From Many Silos at Web-scale Makes E-content Truly  Discoverable
Harvesting From Many Silos at Web-scale Makes E-content Truly  Discoverable
Harvesting From Many Silos at Web-scale Makes E-content Truly  Discoverable
Harvesting From Many Silos at Web-scale Makes E-content Truly  Discoverable
Harvesting From Many Silos at Web-scale Makes E-content Truly  Discoverable
Upcoming SlideShare
Loading in...5
×

Thanks for flagging this SlideShare!

Oops! An error has occurred.

×
Saving this for later? Get the SlideShare app to save on your phone or tablet. Read anywhere, anytime – even offline.
Text the download link to your phone
Standard text messaging rates apply

Harvesting From Many Silos at Web-scale Makes E-content Truly Discoverable

2,433

Published on

For many years, libraries have been dreaming about a simple, easy, fast search solution that unifies all of the resources into a single repository. In the current model, the user is faced with the …

For many years, libraries have been dreaming about a simple, easy, fast search solution that unifies all of the resources into a single repository. In the current model, the user is faced with the problem of dealing with multiple information silos and no-compelling starting place in implementing their search. Recently, the introduction of a “Web-scale discovery” layer, or Next-Generation Catalog, can provide this starting point for library patrons. This talk will discuss how these Next-Generation library discovery applications can go beyond the local library holdings and beyond federated search to offer a single Google like search service across all local and subscription resources.

Published in: Education, Technology
0 Comments
1 Like
Statistics
Notes
  • Be the first to comment

No Downloads
Views
Total Views
2,433
On Slideshare
0
From Embeds
0
Number of Embeds
0
Actions
Shares
0
Downloads
14
Comments
0
Likes
1
Embeds 0
No embeds

Report content
Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
No notes for slide
  • VAC: create nebulous paths for existing products: Voyager, Aleph support will continue, but how much longer will it really be in EL’s interest to continue developing them, given URM? Some niche modules like Media Scheduling have already fallen by the wayside. How much longer will it make sense for us to stick with the current ILS? If libraries offer alternatives to the front-end, they’ll learn lessons about the backend too. The vendors know that in these budget times, we need better back-office efficiency. CLIR’s No Brief Candle report and plenty of other sources say we need to change. sometimes reduce choice as products are merged 360Search + WebFeat = ?? new products evolve rapidly See Primo -> services -> Central insert new consortia/vendors into existing markets Not just for “library automation” vendors anymore – using corporate relationships to leverage content (SerSol and ProQuest) OCLC’s WorldCat Local, now it’s adding articles and Web-scale acq, cat and circ functions One company sticks its neck out and does something innovative, and other companies and industry players then race to catch up to that: SerialsSolutions launched Summon, and now EBSCO, Ex Libris, and OCLC are all chasing after them into that market segment. Open source: Evergreen’s acq/serials not very mature yet, and many academics are hanging back to watch UPEI and Conifer for a while. We have the Michigan Evergreen consortium of public libraries, and the academics have had one conversation so far about it, but that’s all to date. Products years away: OLE’s still to work out funding to have something non-partners can use by mid-2013 (not to mention the tectonic shift that’s going to need to happen for enterprise IT to allow the library to leverage systems like Banner, PeopleSoft, and Blackboard rather than continuing to have a “patron database” in the ILS). Ex Libris just announced URM development partners, so who knows when there’ll be a mature URM that the larger customer base can implement? Flux, uncertainty and disruption force very careful choices in tight budget times – now’s not the time to be changing ILSs or implementing fed search tools, IMHO
  • Nina McHale spoke about this at length yesterday. You can create some metasilos If one silo fails, you may get incomplete results PsycINFO may be fast, but what about other silos that are slower? WebFeat: you can’t do a three-term search with Booleans if all X databases haven’t implemented such functionality the same way It’s a starting point.
  • Early Primo was an overlay that still showed you Voyager records, still ran a db search federated Later Primo started harvesting and using Web services so you didn’t have to see Voy anymore, but Vanderbilt still uses fed search – watch the timebar Now PrimoCentral is a big index that’s Web-scale/cloud based. Growing number of publishers joining. That’s three evolutions libraries and users have had to grok over 2-3 years. The best new systems use tools like Lucene/Solr: Ex Libris, VuFind, etc. Why bother cataloging things in such detail if we can’t use the fields? We need to choose how we index things (though public services and tech services may not agree – the more fields we include, the broader the search), though we need to think carefully about how much time to spend on things that go further and further down the long tail.
  • Early Primo was an overlay that still showed you Voyager records, still ran a db search federated Later Primo started harvesting and using Web services so you didn’t have to see Voy anymore, but Vanderbilt still uses fed search – watch the timebar Now PrimoCentral is a big index that’s Web-scale/cloud based. Growing number of publishers joining. That’s three evolutions libraries and users have had to grok over 2-3 years. The best new systems use tools like Lucene/Solr: Ex Libris, VuFind, etc. Why bother cataloging things in such detail if we can’t use the fields? We need to choose how we index things (though public services and tech services may not agree – the more fields we include, the broader the search), though we need to think carefully about how much time to spend on things that go further and further down the long tail.
  • Transcript

    1. <ul><li>Harvesting from Many Silos at Web-scale Makes eContent Truly Discoverable </li></ul><ul><li>George Boston </li></ul><ul><li>Electronic Resources and Serials Librarian </li></ul><ul><li>Western Michigan University </li></ul><ul><li>Scott Garrison </li></ul><ul><li>Associate Dean for Public Services and Technology </li></ul><ul><li>Western Michigan University </li></ul>ER&L 2010
    2. The Challenge – How Best to Serve the Users… <ul><li>Broad discovery of both known and known items in our collections, not just in their discipline </li></ul><ul><ul><li>Students aren't as concerned where the </li></ul></ul><ul><ul><ul><li>results come from, as long as it's relevant </li></ul></ul></ul><ul><ul><li>Faculty need interdisciplinary research </li></ul></ul><ul><li>Be more like Google: simple, easy, fast </li></ul><ul><ul><li>Fewer places to look for more kinds of content </li></ul></ul><ul><ul><li>Single search interface </li></ul></ul><ul><ul><li>Get to the actual item in fewest clicks possible </li></ul></ul>ER&L 2010
    3. Users are faced with wading through several information silos to get what they need: <ul><li>OPACs </li></ul><ul><li>Electronic resources (e-journals, e-books, databases, articles, etc.) </li></ul><ul><ul><li>Separate interfaces </li></ul></ul><ul><ul><li>Multiple targets for the same resources </li></ul></ul><ul><li>Local Resources (ContentDM, Luna) </li></ul>ER&L 2010
    4. ER&L 2010 <ul><li>Vendor acquisitions, consolidation, catch-up </li></ul><ul><li>Some products are still years away </li></ul><ul><li>Open source options are emerging </li></ul><ul><li>New standards are forming </li></ul><ul><ul><li>FRBR, FRAD, FRSAR, KBART, RDA, etc. </li></ul></ul><ul><li>These lead to great FUD (Flux, Uncertainty, Disruption) </li></ul><ul><ul><li>Everything we're doing is on shifting sands </li></ul></ul>Changing Marketplace:
    5. ER&L 2010 <ul><li>Allows some general, discipline searching </li></ul><ul><li>Mixed, incomplete results </li></ul><ul><ul><li>Especially if one or more databases “fail” </li></ul></ul><ul><li>As slow as the slowest silos </li></ul><ul><li>If local, very network-inefficient </li></ul><ul><li>Many different metadata schemas, less sophisticated searching </li></ul>One solution: Cross-silo Federated search
    6. Another solution: “Web Scale discovery” <ul><li>Metadata is harvested from various content sources and searches are conducted against a single index, this... </li></ul><ul><ul><li>Reduces latency that federated searches have </li></ul></ul><ul><ul><li>Normalizes the metadata in a common search interface for a more consistent search and display </li></ul></ul><ul><ul><li>Easily scalable </li></ul></ul><ul><ul><li>Hardware upgrades to the “Cloud” </li></ul></ul><ul><ul><li>No need to rely on Vendor hardware upgrades </li></ul></ul>ER&L 2010
    7. ER&L 2010 <ul><li>Searching for the 21 st century </li></ul><ul><ul><li>One search box </li></ul></ul><ul><ul><ul><li>Searching through multiple silos at once </li></ul></ul></ul><ul><ul><ul><li>User-centered design, based on research </li></ul></ul></ul><ul><li>Puts our metadata to better use </li></ul><ul><ul><ul><li>Faceted searching </li></ul></ul></ul><ul><ul><ul><li>Relevant results are presented to the User </li></ul></ul></ul>Advantages of Web Scale discovery
    8. ER&L 2010 <ul><li>Built on 21 st century technology </li></ul><ul><ul><li>Ajax, API </li></ul></ul><ul><ul><li>SOLR – provides speed optimized indexing </li></ul></ul><ul><li>Highly configurable interfaces </li></ul><ul><ul><li>Design the interface to look the way we want </li></ul></ul><ul><ul><li>Can integrate well with other discovery tools (VuFind, Primo, etc) </li></ul></ul>Other advantages…
    9. “ Web Discovery” solutions: <ul><li>Serials Solutions Summon </li></ul><ul><li>Ex Libris Primo Central </li></ul><ul><li>Ebsco Discovery Service </li></ul><ul><li>Innovative Interfaces Encore Discovery </li></ul><ul><li>OCLC WorldCat Local </li></ul>ER&L 2010
    10. ER&L 2010 Summon http://wmich.summon.serialssolutions.com/
    11. ER&L 2010 Summon Search Results:
    12. <ul><li>Primo Central </li></ul><ul><li>http://www.exlibrisgroup.com/category/PrimoCentral </li></ul>ER&L 2010
    13. <ul><li>EbscoHost Discovery Service </li></ul><ul><li>http://www.ebscohost.com/discovery/default.php?id=1 </li></ul><ul><li>Demo: </li></ul><ul><li>http://www.ebscohost.com/discovery/flashDemos/CDSDemo/CDS_Demo.htm </li></ul>ER&L 2010
    14. <ul><li>Encore Discovery </li></ul><ul><li>http://encoreforlibraries.com </li></ul>ER&L 2010
    15. <ul><li>WorldCat Local </li></ul><ul><li>http://www.oclc.org/us/en/worldcatlocal/default.htm </li></ul>ER&L 2010
    16. Implementing Summon at WMU <ul><li>Why discovery at WMU? </li></ul><ul><ul><li>Dissatisfaction with federated search options </li></ul></ul><ul><ul><ul><li>Cost was a big factor (MetaLib required Primo – so we'd need two products rather than just one) </li></ul></ul></ul><ul><ul><ul><li>We tried a 360 Search trial – results disappointing </li></ul></ul></ul><ul><ul><li>LibQual survey </li></ul></ul><ul><ul><ul><li>Users wanted a single search box – ala Google </li></ul></ul></ul><ul><ul><li>Strategic plans – University and Library </li></ul></ul><ul><ul><ul><li>Optimizing discoverability of resources </li></ul></ul></ul>ER&L 2010
    17. WMU background <ul><li>Carnegie research university </li></ul><ul><li>Undergrad enrollment = 25,000+ </li></ul><ul><li>5 libraries serving multiple sites statewide </li></ul><ul><li>1.7M+ bibliographic records </li></ul><ul><li>400+ databases </li></ul><ul><li>4,500+ print journals </li></ul><ul><li>42,000+ online journals, newspapers </li></ul><ul><li>80,000+ online books </li></ul><ul><li>Voyager, SFX, EZproxy, CONTENTdm, Luna </li></ul><ul><li>Summon and VuFind as discovery systems </li></ul>ER&L 2010
    18. Timeline for implementation <ul><li>2009 March – signed as Beta partner </li></ul><ul><li>2009 June – Initially operational as a “pre-Beta” </li></ul><ul><li>2009 September – Rolled out to the public as a “trial” </li></ul><ul><li>2010 January – Integrated into the web site </li></ul><ul><li>2010 Fall – Fully integrated into main tabbed box </li></ul>ER&L 2010
    19. Integrating Summon into our web site ER&L 2010
    20. Set up challenges <ul><li>Extract local data from OPAC (Voyager) </li></ul><ul><ul><li>Previously this had been done for VuFind, and </li></ul></ul><ul><ul><li>MelCat (a Michigan-wide Union catalog) </li></ul></ul><ul><li>Extract From our ERMS (Verde) </li></ul><ul><li>Extract from SFX </li></ul><ul><li>Tweak information in Summon client center </li></ul>ER&L 2010
    21. Configuration challenges <ul><li>Consistent metadata </li></ul><ul><ul><li>Backstage clean-up of OPAC data </li></ul></ul><ul><ul><ul><li>Requires Summon to re-ingest/re-index the catalog data </li></ul></ul></ul><ul><ul><li>“ Match and Merge” to eliminate duplicate entries for articles – this is still in development </li></ul></ul><ul><li>Inconsistent linking to resources </li></ul><ul><ul><li>Only good as publisher metadata and interface </li></ul></ul><ul><ul><li>Not all vendors are OpenURL “compliant” </li></ul></ul>ER&L 2010
    22. On-Going maintenance challenges <ul><li>Summon Client Center is a bit clunky </li></ul><ul><ul><li>Still a work-in-progress </li></ul></ul><ul><li>Limited usage statistics capability </li></ul><ul><li>No tool for local configuration in the client center yet </li></ul><ul><li>Full integration into Client Center expected later in 2010 </li></ul>ER&L 2010
    23. Using the system <ul><li>Advantages </li></ul><ul><ul><li>Agile development cycle – frequent feature releases </li></ul></ul><ul><ul><li>Integrates well with VuFind </li></ul></ul><ul><ul><li>Good place to begin research </li></ul></ul><ul><ul><li>Great for interdisciplinary research </li></ul></ul><ul><li>Disadvantages </li></ul><ul><ul><li>Can't easily specify targeted index searching like in VuFind – Summon basically searches one big index </li></ul></ul><ul><ul><li>It's not built like the search interfaces we've used in the past </li></ul></ul><ul><ul><li>Hard to tell what exactly is included – not a clean match </li></ul></ul>ER&L 2010
    24. Evaluation and the way forward <ul><li>Focus groups and Usability Studies </li></ul><ul><li>“ What databases are included?” </li></ul><ul><li>Better interface design </li></ul><ul><ul><li>Use of the API allows us to build an interface </li></ul></ul><ul><li>Integrating into Subject and other guides </li></ul><ul><li>Will it allow Librarians to stop teaching about interfaces, but rather how to evaluate results? </li></ul>ER&L 2010
    25. <ul><li>Thank you for your attention. </li></ul><ul><li>Questions? </li></ul><ul><li>George Boston - [email_address] </li></ul><ul><li>Scott Garrison - scott.garrison@wmich.edu </li></ul>ER&L 2010

    ×