ALA TechSource Logo


curve Home spacer Publications spacer Subscribe spacer Blog spacer About  

Building Comprehensive Resource Discovery Platforms

Submitted by Marshall Breeding on March 4, 2011 - 8:56am

This column appears in the March 2011 issue of Smart Libraries Newsletter. To read more from Marshall Breeding on mobile library technology and other facets of the library automation industry, you can purchase this issue or subscribe to Smart Libraries Newsletter at our metapress site.

One major trend in the library automation industry that I have been following in recent years is the evolution of discovery systems toward a model that provides access to a more comprehensive representation of library collections. A variety of products and projects have emerged with the ambitious goal of addressing the content represented in a library’s subscriptions to electronic resources in addition to the books and other materials managed in the integrated library system.

One of the strategies behind these products is the creation of massive consolidated indexes created out of citation metadata or the full text of articles harvested from the publishers and providers of content to libraries. This model of discovery creates an index spanningall types of library content that can provide fast retrieval of search results. With articles represented in the index, users can more easily discover items of interest to their research, and click through from the search results to the electronic article on the provider’s server. By making metadata or full text available to the discovery service, a publisher of electronic content gains better exposure of their content, while retaining control over the display or delivery of that content.This model of discovery depends on the cooperation between the producers of content and the creators of discovery services. Some of the major discovery providers that follow this approach include the Serials Solutions’ Summon service, EBSCO Discovery Service, OCLC WorldCat Local, and Ex Libris Primo Central. Each of these organizations has aggressively pursued content providers to contribute data that can be indexed within their discovery service to help achieve the maximum degree of comrehensive coverage of the content to which libraries subscribe.

An event that recently took place in this arena involved EBSCO’s move to discontinue making its content available to Ex Libris for inclusion in Primo Central. EBSCO originally made an agreement to provide citation data from its EBSCOhost products to Ex Libris in July 2009, prior to releasing its EBSCO Discovery Service, a direct competitor to Primo Central. The competitive issues between their two discovery products outweighed any advantages EBSCO would have by cooperating with Ex Libris to contribute content to enhance Primo Central.

While losing the EBSCOhost content in Primo Central is a short-term disruption, it is not necessarily a long-term obstacle to the development of Primo Central in particular or a general mark against this model of discovery based on large aggregate indexes. When populating an index for a discovery service, it’s helpful to gain access to large numbers of articles in a single package through a deal with an aggregator. It’s also possible to represent the same content in the indexes through cooperative arrangements with the primary publishers and content providers covered within any given aggregated database of article content. Discovery service developers do well to pursue multiple paths in parallel to ensure the maximum coverage of articles and to hedge their bets relative to competitive issues such as the one between Ex Libris and EBSCO.

To the extent that libraries favor this approach of discovery based on consolidated indexes, they have an interest in the highest level of cooperation between the publishers and providers from which they license content and the organizations that offer discovery systems. When a content provider opts out of making their materials available to discovery products for indexing, it causes problems for libraries that depend on discovery services. I believe that it’s mutually advantageous to both publishers and discovery providers to cooperate, since it both increases the effectiveness of the discovery products and improves the value of the content for ibraries as it makes that content more easily available to their users.

Libraries can take part in the process of increasing the content available within discovery services by bringing this issue into their selection and procurement process. In the same way that libraries routinely require license terms for providers of content products support practices such as COUNTER statistics for measuring the use of materials, SUSHI for automatically delivering those statistics, or OpenURL for linking, it would likewise be reasonable for libraries to introduce requirements that vendors make content available to the discovery services provider of their choice for the sole purpose of indexing. Although there seems to be a broader acceptance of content providers to work with discovery systems, making it part of the license terms will help close the gap on the content not currently supported in this important genre of library software.

Comments (2)

Marshall--I agree with the

Marshall--I agree with the previous commenter that you have raised an important point. For those of us that work at libraries that have implemented a discovery layer, we should definitely be advocating that the content vendors we work with make their content available for indexing by the discovery tool; it’s in everyone’s interest for that to happen. As to whether we should, as you suggest, make it a license requirement, I think that might be a bridge too far: it would in many cases unnecessarily delay access to users and draw out the acquisition process. But, again, I agree with your overall point.

Beyond advocating that content providers allow their data to be indexed, we should also be advocating that access platforms work with the vendors of discovery layers to develop better linking between result listings in discovery tools and full-text content on the access platform. Indeed, one of the biggest problems I see with discovery layers at present is that too often, a result listing on a discovery layer links incorrectly to the full-text on an access platform.

I think that the previous commenter raises an important point regarding entanglements between accuracy of search results and the interests that the vendors of discovery layers have in increasing use of their products. For more discussion on that point, see the following posting on Against the Grain Multi-Grain forum:


You raise an important point

You raise an important point (I like the idea of trying to secure index neutrality in contract discussions) but I'd like to see you dig a bit deeper into the conflicts of interest that define this particular library tech niche.

Several "discovery" system vendors are either engaged in, or owned by, corporate entities that also provide content (e.g., Summon with ProQuest, EBSCO with, well, EBSCO). Can we just assume that a company that has a vested interest in promoting their own content won't also skew their "discovery" results so that their content is featured?

Is it really "discovery" when you're just poking around one vendor or another's walled garden of information?