Serials Solutions announced a new discovery service called Summon that aims to provide a unified search tool for library users. The goal of this tool is to provide access to all of a library’s content, print and electronic, through a single search box.
At the heart of summon lies a very large index of content that the company is building from a variety of sources. This index includes the full text of a large body of articles provided by many different publishers as well as open access content. Summon aims to expand its index to include all of the content represented in a library’s collection of subscriptions, its local print holdings and other resources.
Fueled by content from publishers and open access ProQuest and Gale are already moving content into the Summon index. Others participating include Springer, Taylor & Francis, Sage publications, Nature Publishing Group, Cambridge University Press and Oxford University Press, with others likely to join in the near future. It’s not surprising that ProQuest, the parent company of Serials Solutions, would provide content for this product. Seeing companies like Gale, a publisher that competes with ProQuest, providing content reflects an important step in the industry where publishers are willing to provide copies of their content to a discovery tool.
During the initial beta test period, Serials Solutions indicated that Summon already indexes over 300 million items representing 50,000 journal titles from more than 40 publishers. Although the content included during the beta stage is impressive, it does not yet cover the complete universe of material represented in library subscriptions. Serials Solutions aims to attract a very broad base of publishers to provide their content for Summon. The fundamental idea of Summon involves harvesting data from publishers to power its discovery engine, but once users select an item, they are directed to the content on the publisher’s site. If this model of discovery proves effective, it seems reasonable that providing the full content ultimately increases the use of the material from the publisher, which reinforces its value. This increase in use and value should offset any risks that publishers take in releasing copies of their strategic assets for use in a discovery platform.
Beyond Federated Search
This model of consolidated searching based on content harvested in advance is a great improvement over federated search technique, where an intermediate interface performs real-time searching against multiple information resources. The distributed query model of federated search is constrained by many factors--practical limits on the number of targets simultaneously searched and the amount of records returned from each target. Search operations done in this way can be far from instantaneous, and often produce unsatisfying results.
Federated search tends to rely on protocols such as Z39.50 and XML gateways to communicate with remote information resources. This model depends on connectors that must be built and maintained for each remote resource. They must contain the specifications on interaction with the federated search engine, details of the protocols used, the structure of how data are returned, and on other details needed to extract and present results from that target. These connectors tend to be fragile—any changes that an information publisher makes in their service can break them.
The modern technique for searching diverse information resources involves harvesting content from each of the resources and building a centralized index. This approach requires much more work on the front end to gather comprehensive sets of metadata and to build the index, but by performing these procedures in advance, users gain the ability to search almost instantly.
It also eliminates the problem of shallow results from each target since a centralized index can process very large numbers of results and more effectively perform relevance weighting, cluster results into sets, present facets, and deliver many other features expected in modern discovery environments.
The distributed query model of federated search can place a tremendous burden on the publisher’s technical infrastructure. These federated search platforms can dramatically increase the number of searches performed and the number of records returned—most of which are never seen by an end user. By offering their content for ingestion into a discovery engine, publishers can eventually ease some of the load imposed by the older style of real-time distributed query federated searching.
Summon goes beyond searching of citation-level metadata on the articles in its index, delivering search results based on the full text of the articles. This foray into full-text searching across content from many different publishers represents a major advancement in the capability of discovery products.
Print and multimedia content included Summon aims to provide access to all of the library’s resources, print or electronic. One component of Summon involves harvesting all of the records from the library’s integrated library system and loading them into its index, enabling books and articles to be searched simultaneously. As summon presents information about items from the library’s print collection, it will retrieve holdings and availability information from the ILS in real time, as do many of the other discovery products. This approach can also be extended to multimedia content maintained by the library in other local repositories.
Offered through Softwares-a-Service
Summon will be delivered as a hosted service. Serials Solutions embraces Software-as- a-service as its strategic approach. Libraries will not have to deal with the creation and maintenance of the very large-scale indexes that power Summon. Serials Solutions aims to extend Summon to include a very broad range of information resources. It’s critical that search results correspond to the library’s subscriptions. The configuration of Summon will involve establishing the profile of the library’s subscriptions and the individual titles represented within them in much the same way this process is done with the company’s 360 Link product.
Next-gen Interface features
The interface provided with Summon includes many of the features that have become expected in the new wave of discovery interfaces. Some of the main characteristics include a single search query box, relevancy ranked results, filtering, search suggestions (“did you mean?”), and faceted navigation.
Designed for integration
In addition to the direct end-user interface, Summon will also offer an application programming interface, or API, that allows the library to make the search capabilities of Summon available to other components in its technical environment such as courseware systems and other next-generation interfaces. The Summon API, for example, can also be used in conjunction with interfaces such as AquaBrowser to blend its interface with a more expansive search environment.
Summon was first announced at the ALA Midwinter Meeting in Denver, CO in January 2009. Two libraries, Oklahoma State University and Dartmouth College, have been involved as development partners since October 2008. Additional beta sites will join through the general release of the service planned for about July 2009.The product primarily targets academic libraries. Serials Solutions has not yet described pricing options for Summon.
A formidable Competitor among Next-Gen interfaces
Summon embodies many of the ideal features of a next generation library interface. Through a design that includes the ability to incorporate the content of the local ILS and full text or citation representations of the library’s electronic content it aims to deliver a single point of entry into all of the content offered by a library. Summon challenges existing products in the discovery interface arena including Primo from Ex Libris, Encore from Innovative Interfaces, and open source interfaces such as VUFind initially developed at Villanova University. While Summon comes in as a relative latecomer in this product genre, its massive index of content will make it an attractive consideration for many academic libraries in an arena where the interface features have become less differentiated. Summon represents a bold next step in the development of an arsenal of products that Serials Solutions offers to academic libraries to help them manage and provide access to their content.