What are the most common reasons openurl linking fails?

What are the most common reasons openurl linking fails? How does one improve chances of openurl linking working?

aarontay

• e-resources
• openurl

Comments

• Peter Murray: Two NISO projects will probably be of interest: IOTA (Improving OpenURLs Through Analytics - http://www.niso.org/workrooms/openurlquality ) and KBART (Knowledge Base And Related Tools - http://www.niso.org/workrooms/kbart ).

Answer by phette23

I'm not terribly experienced with openURL but I'll take a shot at this:

• Configuration, configuration, configuration. You need to make sure the parameters for each resolver link match up with what actually goes on in the database; if a parameter name has changed or been removed in a target resource, then the link will crash even if the full-text really is available. I saw this just yesterday with EBSCO LinkSource -> Literature Online (LION).
• Encoding issues. A string might run through three different web forms (original search form -> link resolver -> proxy server) which gives ample opportunity for an encoding mistake, either because of mismatched types (e.g. UTF-8 vs. Windows-1252 vs. whathaveyou) or improper encoding functions (e.g. escape() vs. encodeURI() vs. encodeURIComponent() vs. some other broken encoding function in a poorly written library).
• Inaccurate knowledge base. Your knowledge base says a source holds the full text for a particular journal when in fact it does not.

That's a lot to track down but it should be fairly obvious given a test case. You just need to see what kind of error results, like playing telephone in reverse.

Comments

• dsalo: Good list. Lousy vendor metadata can sometimes play a role as well; also the general difficulty of matching correctly on certain fields (such as uncontrolled author names).

Answer by Jakob

Maybe this is not the most common reasons why OpenURL linking fails, but the inaccessibility of knowledge bases is a showstopper of this technology. Most data about which journals are licensed by which library is managed in closed silos. Moreover the knowledge base is often inaccurate because libraries do not update it. I doubt that this will change unless knowledge bases only aim at users. If a librarian looks up his own holdings, the information does not come from the knowledge base, does it? If errors in the knowledge base would affect the daily work (and budget) in the library, the accuracy would be better.

Comments

Answer by Jaclyn

There are so many potential points of failure that it's often difficult to pinpoint the cause. That said, Cindi Trainor and Jason Price took a stab at it (Trainor, Cindi; Price, Jason. Library Technology Reports, Oct2010, Vol. 46 Issue 7, p15-26), but their results were pretty evenly spread out. As to what you can do: (1) keep your part of the kb as up to date as possible; (2) encourage database/journal vendors to provide accurate data to the knowledge base providers; (3) encourage vendors of all types to adhere to standards (are they OpenURL compliant?) and contribute to initiatives like KBART that aim to improve knowledge bases & linking. It also wouldn't hurt to look at your failures within your link resolver to see what's commonly not working, and see if you can fix it/encourage your link resolver/kb provider to fix it. [We recently did this with linking to dissertations, for example].

Comments

• Joe: Trainor & Price also had a shorter article in American Libraries which is open access : Rethinking Library Linking: Making OpenURL Better with Data, Data, and More Data.