Google and OpenURL resolvers

2005 June 13
by Karen

I’ve been following discussion about Google adding links to a library’s OpenURL resolver in a number of places including the SFX listserv, Web4Lib, and Making Links and anticipating our library’s migration to version 3 of SFX so that we can have links from Google Scholar to our OpenURL resolver. Now that we have a version 3 instance of SFX, I can start thinking about this more seriously. Honestly, I am both fascinating with the prospect and slightly skeptical about the process. Particularly since we have to export our holdings in order to get the links in Google Scholar.

Because of I still really haven’t decided how keen I am on this whole idea of linking Google Scholar and libraries, I’ve been dilgently reading everything I can find about this topic . The Digital Librarian has a good post about Google Scholar’s terms of service. Gary Price has a profile of the service which allows libraries to add OpenURL links and the Distant Librarian has commentary on Google Scholar developments. There seems to be many potential positive possibilities. However, there also appears to be negative consequences as well.

One concern I have about Google Scholar and OpenURL linking is that Google seems to be under the misapprehension that OpenURL is all about linking to fulltext. This is far from the case. OpenURL can be used to link to a variety of things. It can be used to submit a request to an Interlibrary Loan System or to add citations to a citation management tool. It can also link users to unique collections that a library owns. I think it is exceptionally shortsighted for Google to be focusing on this technology only as a link to fulltext. It is like saying to OCLC “we aren’t interested in your WorldCat records that are for print materials”. Whether we like it or not, not everything is digital. OpenURL is a great tool to help users figure out where they can get a copy, physical or digital. It can possibly reduce the number of clicks a user has to make to figure out if the book (found via OpenWorldCat) is at their library. Particularly since OpenWorldCat can’t tell what your library is unless you are on campus. Furthermore, using it just to link to fulltext is an exercise in format discrimination and limits what resources users have at their disposal. If users want to limit by format, let them don’t choose to do so for them.

Another limitation of Google Scholar is the fact that it relies on IP address to determine what your institution is (and therefore where your OpenURL resolver is located). So if I am a SUNY Cortland student using computers off campus I won’t get the OpenURL links. At least based on what I understand of the technology behind the scenes I don’t think I’ll get the links. This is too bad because the people who most often need to be directed to library e-resources are accessing materials from off-campus (distance students in particular). The only way around this issue seems to be having users install some software like the Firefox extension that adds OpenURL links. Things like Wag the Monkey also have possibilities. However, even these option don’t solve the more over arching problem that users have access to many collections (both physical and electronic) through their libraries (academic, public, etc). Somehow libraries need to figure out a way to allow users to interact with all the collections at their disposal without having to know where these collections are virtual or physically. In the meantime, OpenURL linking in GoogleScholar is a nice step forward that is at least worth investigating to see how it works and how well it works. I’ll keep everyone up to date on my progress getting it working at Cortland.

3 Responses leave one →
  1. 2005 June 13

    I have just a couple of comments. For off-campus access, students may register with Google Scholar and specify the institution(s) for which they are affiliated. This is done via the Scholar Preferences ( under ‘Institutional Access’. It’s not as seamless as the IP-based authentication, but it does provide a quick setup mechanism to make use of your link server.

    Also, Google Scholar does provide both the capability to link to the full-text as well as to access the full SFX menu. For example, I defined Stanford as my library and searched on ‘asthma cure’. By clicking on the text link ‘Find It @ Stanford’ I can access SFX and all available services.

    Finally, if you do decide to participate with Google Scholar (after having decided to export your holdings), SFX provides an easy mechanism to export your holdings information as an XML file which is then harvested by Google (the process of registering with Google as an institution – also handled by SFX – provides the appropriate location information for the XML output file).

  2. 2005 June 14
    Administrator permalink

    Isn’t it the case though that the SFX link only shows if fulltext is present? If so then, if we don’t have fulltext then the link won’t show so users can use the other services that are part of SFX.


  3. 2005 June 14

    The SFX link appears for every citation. It appears more prominently (next to the title) for items your library holds in fulltext (assuming you’ve made your holdings available to Google to crawl).

    Also, technically, WAG the Dog/Monkey/Menagerie is perfectly happy to find print items from multiple libraries for the user. If the item is a journal, it defers to SFX to handle this, but if it runs across an ISBN, it searches any number of catalogs that are associated with a particular institution. For instance, Georgia Tech’s WAGger searches GT’s catalog and, if it finds no matches, the state union catalog. For the WAGNet (which takes your result set and looks for more related items based on LCSH and database coverage), it searches the local catalog, union catalog and Library of Congress catalog.

    But, you know, we can promote ILL and document delivery and whatnot all we want… the user is much more concerned about fulltext instant gratification. All the user training in the world isn’t going to change that.

Leave a Reply

Note: You can use basic XHTML in your comments. Your email address will never be published.

Subscribe to this comment feed via RSS