IL2004 - Monday - 13:15 Federated Searching and OpenURL (Part I)

Originally posted 2004-11-16.

“An Introduction to Federated Searching and OpenURL”
Frank Cervone
Assistant University Librarian for Information Technology
Northwestern University
Evanston, IL, USA

[Oh I see THEY have Internet (the presenters)]

federated search - search multiple resources (in background) - provide unifed results list

using OpenURL to link to full text in results list

“the vast majority of OpenURL is done using URL rather than DOI”

presents OpenURL as a way to link source database to target database, e.g. from FirstSearch get OpenURL, your resolver turns that into a linkt to EBSCO

this is as opposed to having direct, internal links (typically to internal resources) in the source database

OpenURL
- redirect through resolver
- resolver interprets data which may be
-- DOI
-- a URL encoded with metadata about the resource
- locates appropriate copies

they implemented SFX, it took them about 3 months
but Northwestern had “serial solutions” data
another place didn’t have that data and it took them 2 years to key it all in

they prioritized their results based on availability, comprehensiveness, whether you actually get it consistently when you click on it, quality of PDF etc.

you then have on-going maintenance of your OpenURL database, but most of this should be automated

“as nice as OpenURL is, it is just a stop on the destination to federated searching”

what is federated searching
- metasearch
-- uses metadata to make decisions
- megasearch
-- uses full text to make decisions
-- e.g. dogpile, alltheweb

federated search engine - ezproxy (remote auth and acces) - world of elec tronic resources

once you have located the info (using federated searching) then use OpenURL to locate appropriate copy

people overwhelmingly choose a single, common, unified interface

they are building federated search for undergrads
live search
shows # of hits
can look at individual result sets

users (e.g. faculty) can select which databases they want
they have set a limit of 8 databases that you can federate at a time

also have “my resource list” - user’’s last search, favourite resources (databases)
as well “my e-journals list” which shows the journals covered by the fave resources

OpenURL steps...
- linking to providers - get them to enable OpenURL, interfaces etc.

federated search major issues
- time to cnfigure databases and resources
- dedup results / relevance ranking
- defining searchable collections

They are using ExLibris for federated searching.

“The Long and Winding Road: Evolving E-Journal Management and Discovery Tools”
Cindi Trainor
Director, Information Technology
The Libraries of the Claremont Colleges

they use lots of Serials Solutions stuff
they also got RefWorks

doesn’t seem to actually be anything about federated search and OpenURL

she just gave us her ezproxy access

Comments

[cindi]

Thanks for posting the note about EZproxy access. We had made [a poor choice of passwords]. Uh, we're changing that... :) I trust that you only used your power for good.

[Edit by Science Library Pad, 2004-12-30]