:: Digital Libraries Columns


Library Journal "Digital Libraries" Columns 1997-2007, Roy Tennant

Please note: these columns are an archive of my Library Journal column from 1997-2007. They have not been altered in content, so please keep in mind some of this content will be out of date. :: Digital Libraries Columns

Linking, Not Thinking


   Though open URL resolvers are a major step forward in guiding users to
   the right copy of the item they seek, they are not the complete
   solution. Instead of putting yet another window up in front of the
   hapless user that is often chock-full of confusing choices, we need to
   convey users automatically to the full text if it's available, or to
   the next best thing. In other words, we should view most link resolver
   windows as a failure to give users what they want. I bemoaned the
   primitive state of OpenURL linking services in a [145]recent column (LJ
   7/06, p. 28), but, thankfully, this is beginning to change.

   Enter the Umlaut

   One sign of better linking comes from Ross Singer of Georgia Tech.
   Singer has put together a linking system, the Umlaut, that just won 
the OCLC Software Contest by beating out 11 other entries.

   The Umlaut is a set of software routines that uses several web 
   to create more effective links to content as well as an OpenURL
   resolver. Singer explains that when an item is requested, "the Umlaut
   takes the user's IP address and asks the OCLC Resolver Registry if
   there are other link resolver services associated with this IP. If the
   response is yes, the Umlaut adds those services to the user's
   collection." This provides the widest range of OpenURL services to a
   given user (for example, when a user is served by both a local library
   and a consortium of which that library is a part).

   The Umlaut then tries to enrich the item metadata by querying
   additional sources of information depending on what the original
   request includes. For example, if there is a digital object identifier
   (DOI), CrossRef is queried for additional metadata. PubMed IDs are also
   used in this fashion, and the architecture of the application can
   additionally handle other identifiers (e.g., Handles). After trying to
   enrich the item metadata, the Umlaut queries the OpenURL resolver, 
   local and regional library catalogs. If the item has an ISBN, the
   Umlaut uses OCLC's xISBN service to retrieve all related ISBNs with
   which it queries the catalogs.

   Conference proceedings are tricky for linking, so Singer extracted data
   from the library catalog and indexed fields useful for finding them.
   The Umlaut also queries and Citebase for preprints and
   retrieves book cover images from The display is much easier
   to use than any OpenURL resolvers I've seen.

   Getting users to full text

   The University of Rochester (UR) Libraries, NY, has been working to
   create effective user-centered finding services like the Getting Users
   to Full-Text (GUF) project since 2004. Aiming to increase the number of
   links to full text, as well as to provide the next best option, the
   service uses many of the same strategies as the Umlaut does, such as 
   database of journal holdings, OCLC's xISBN service, and the CrossRef
   database, among others. By integrating linking services directly into
   the search results from a metasearch application, UR has been able to
   reduce the click path to two clicks in many cases (one for search, one
   to get full text).

   NextGen linking

   The best linking service is invisible. We rarely need to send users to
   a separate screen, as most OpenURL resolvers do, and force them to
   puzzle over a plethora of choices. We simply need to get them to where
   they expect to be able to go--the item itself.

   Some have objected to this model of not letting the OpenURL resolver
   put up its own window because some libraries like the idea of referring
   to other services like online reference. If libraries survey their
   users they will probably find that no one wants to follow a link to a
   reference service when they really want the item itself. I would like
   to see some data on how many links go bad before making everyone go
   through an often unjustified step.

   Both next-generation linking services are all about intelligent
   integration. Both use a wide array of services provided by other
   organizations, including local holdings information in order to create
   the best possible link. Many librarians may have thought that by
   installing an OpenURL resolver they had "solved" the linking issue. We
   have further to go to do linking well, and, luckily, there are those
   who are showing us the way if we will but pay attention.


                                LINK LIST
   OCLC Resolver Registry
   urlresolver.htm OCLC Software Contest
   announcements/2006-09-28.htm The Umlaut Documentation