OpenURL and EndNote

Contents:

• Reference Linking: Before OpenURL
• Reference Linking using OpenURL
• What is OpenURL?
• How does a Link Resolver work?
• Configuring EndNote to generate OpenURLs
• Using EndNote with a Link Resolver
• Product listing

For some years now, many database vendors have given customers the option of configuring databases so that users can check to see if their library holds a particular reference. This is usually done by generating a search of the library catalogue using the ISSN contained in the reference.

Here is an example of this procedure, using the Library Literature database on the Webspirs platform.

Step 1

Locate a reference to an article which you wish to read. For example, look at record number 5 below. To find out if the library holds this item, click on the "Check for holdings" button.

Step 2

The database sends an ISSN search to the library catalogue and displays the record for the journal.

You might have to go back to the database to check the year of publication of the article, to make sure it is available electronically. Then click on the link for the Emerald full-text collection.

Step 3

This will display a list of all issues available in full-text from Emerald.

Again, you might have to go back to the database to check which issue you require. Then click on the link for the required issue.

Step 4

Now locate the required article. Click on the "View" button.

Step 5

This displays a summary screen, from which you must choose a download format. Request the document in PDF format.

Step 6

The full text of the article will be retrieved.

This procedure requires at least five clicks, and possibly more, if you have to return to the database to check the year and issue required.

2. Reference Linking using OpenURL

If the library has implemented a Link Resolver, it can use the OpenURL protocol to speed up the linking process.

Here is an example showing how this works in a library which has implemented the SFX Link Resolver from ExLibris. We are searching the CINAHL database on the Ovid platform.

Step 1

Reference number 6 looks interesting. To check for full-text, click on the "SFX" button.

Step 2

This connects directly to the OUP journals collection.

Because the library has a subscription to this journal via the OUP Journals collection, you can now click on the link "PDF Version of this Article".

Step 3

This brings up a confirmation screen. Click on the link "Begin manual download"

Step 4

The full text of the article will be retrieved.

This procedure requires a total of three clicks to link from database to full-text.

OpenURL is an Internet protocol, based on a standard maintained by NISO, the National Information Standards Organization.

An OpenURL is a standardized format (syntax) for transferring bibliographic data (metadata) in a URL. The data usually describes an article or a journal. In other words, OpenURL is a way of organizing data in a URL and pointing it to a destination.

Let's see an example of an OpenURL. When searching a database, we might find a reference, as shown below. This database is "OpenURL enabled", and the reference contains a link to "UQ Library holdings". When you click on this link, the database generates an OpenURL.

The OpenURL which it generates looks like this (the colours have been added to identify the different elements):

An OpenURL is made up of three elements:

1. The address of the link resolver (in this case, a WebBridge server at the University of Queensland Library).
2. The source database (in this case, America History and Life from ABC-CLIO).
3. The metadata describing the reference, using the standard OpenURL syntax and field names.

So the OpenURL is a command which is sent to a specific server (a link resolver), and it contains a bibliographic reference which the server has to check. The aim is to locate a full-text copy of the reference, although a link resolver can also perform other types of search.

A link resolver is a web server which contains information on all the electronic collections to which a library subscribes. It decodes the bibliographic data contained in the OpenURL, searches to locate a full-text copy of the reference, and then generates a link to the full-text (if found). It can also search library catalogues to display records for print holdings. It may also offer to submit an inter-library loan request if the item is not located.

The following diagram (supplied by courtesy of Serials Solutions Inc.) illustrates the process.

The process works well when:

• Data providers (databases) are "OpenURL enabled", i.e. they can generate OpenURLs for the references which they contain.
• The link resolver has complete and up-to-date information on ejournal holdings of the library.
• Content providers (full-text collections) support linking, preferably to article level. Some content providers will only link to journal level, so the user still has to navigate to the issue and article required.

As of version 7, EndNote is "OpenURL enabled". However you have to activate this function the first time that you use it. It remains activated thereafter.

Click on Edit|Preferences, then select OpenURL.

Click the box labelled Enable OpenURL. The greyed out text will now appear normally.

By default, EndNote is using a demonstration link resolver on its own website, called 1Cate. If your institution has its own link resolver, edit the OpenURL Path to insert the address of your link resolver.

The Arguments box may need minor amendment to achieve better results with your link resolver.

Click on the OK button to confirm the changes.

To check for full-text, you must first open a reference in your EndNote library.

Then click on References|OpenURL Link.

Your link resolver will either connect directly to the full-text of the article, or report back to tell you what it has found.

In the case illustrated above, click on "Links to Full Text for this journal". If your institution is authenticated to access this resource, the full text of the article will be displayed.

Problems can arise if the reference in the EndNote library contains misspellings or other errors. Abbreviated journal titles and parenthetical additions to journal titles (e.g. Biochemistry (New York)) will cause problems. The presence of an ISSN in the EndNote reference will facilitate OpenURL linking, and for this reason some libraries are checking and amending their EndNote filters and connection files to ensure that they will import the ISSN field.

You can use the demonstration link resolver at ISI (as shown above), and you can even use link resolvers at other libraries. But you will only get consistently reliable results if your library has its own link resolver, which has accurate details of the ejournals to which you have access.

The following are the main link resolvers currently on the market:

• SFX (from ExLibris Ltd.)
• WebBridge (from Innovative Interfaces Inc.)
• VLink (from Geac)
• Horizon Link Resolver (from Dynix)
• LinkSource (from Ebsco Publishing)
• LinkFinder Plus (from Endeavour Information Systems)
• Article Linker (from Serials Solutions Inc.)
• 1Cate (from Openly Informatics Inc.)

Libraries tend to purchase a link resolver from the company who supply their library management system. However there is no obligation to do so. For example, ExLibris claim that 60% of customers using their SFX link resolver are not users of the ExLibris library management system.