OpenURL and cross-searching
I am looking through the paper index to the London Times newspaper to gather information about the Tuskar Rock airplane disaster in March 1968. I find a reference and my heart quickens as the details of this tragedy, related to me on a few solemn occasions by my father, become more real than ever. I approach the librarian at the enquiry desk and hand her the piece of paper on which I have written the reference; she asks me to wait, then returns from the store with the relevant microfilm spool. I sit at the reader machine, find the correct month, then date, then page, then column and there it is. The Aer Lingus Viscount aircraft was flying from Cork to London in good conditions when inexplicably it disappeared from radar and crashed into the Irish Sea by the Tuskar Rock, off the coast of Wexford. The mysterious disappearance was immediately fuelled by claims from local fishermen that the plane had been shot out of the sky by a missile launched from a submarine. This was refuted by the Royal Navy and to this day the cause of the crash that took 61 lives, including my second cousin, the pilot, has remained unknown.
Searching indexes and finding full-text articles has in recent times become a whole lot easier. What appears to be one of the most useful and inspired pieces of thinking has resulted in the OpenURL technology. Many will be familiar with the scenario, quite possibly the current reality for your library; you have hundreds of subscriptions to full text electronic resources; supplemented by abstracts and indexes yet users claim they cannot find or get at the information they require. Why is this? It is because there is no master list or link between the many commercial indexing or abstracting services and, the full text article databases. One way to understand the OpenURL technology is to imagine it as a key part of a bridge that connects the reference or citation in the index to the full-text article. The bridge is known in the parlance as the link resolver or the link server. Either way it carries out the same function; it looks at the metadata contained in the OpenURL of the citation and resolves this to the full text article database. The key point, however, is that it does this within a local context. What this means is that library staff can apply certain parameters to the link resolver such as deciding to which full text databases the user is allowed to link. An example: supposing a user carries out a subject search for “lactobacillus acidophilus”, this may throw up some hits in the life sciences area. Some of the references could be to articles available free on the web, some to articles in a full text database to which the institution subscribes, others to articles in databases which are inaccessible to the user. The link resolver allows the librarian to stipulate in advance how these references will be returned on the hit list; for instance duplicates may be eliminated. Also it provides the opportunity for the librarian to direct the user to the appropriate resource.
It is important to understand that the resolver links are not static; they are defined by the librarian and based on accessibility and authentication criteria. The OpenURL is in fact a string of metadata (usually containing ISBN/ISSN, author, title, journal title, issue number, date and page number)that allows the resolver to provide context-sensitive services. More advanced link resolvers can be embedded into library management systems and so include OPACs and other resources. In this way the example above could also provide a link to the OPAC where a print version of the article is identified. In short the user performs one search to interrogate many sources and is then presented(subject to the parameters set by the librarian) with the most suitable material.
The OpenURL is a standardized way to store and communicate metadata between resources. There are other developments within this technology that are significant and important to its future success, namely digital object identifier (DOI) and CrossRef. DOI is a persistent identifier for intellectual property and evolved from the publishing industry. DOI can be used in conjunction with OpenURL and this provides a powerful combination. CrossRef is more specific to the publisher and allows researchers to link from the citation to the full text article within a publisher’s web site.
One cannot help feeling a little excited about this emerging technology. While search engines such as Google continue to expand their coverage of the surface web, the “deep” or invisible web, which includes all the stuff buried in databases, has up to now remained irretrievable. The OpenURL will change this and have a profound impact on reference services.
Rónán O’BeirneInternet Editor, Reference Reviews and Principal Libraries Officer – Information, Bradford Libraries, Archives and Information, Bradford, UK
