The latest library trend is based on the need to form partnerships to share resources. Libraries must provide ever-increasing services without increasing their budgets or staff. We’re challenged to deliver new suites of digital resources, as well as maintain our traditional operations and services.
In the digital age, the need to function efficiently, find cooperative arrangements, distribute services, and share resources has become an important component of the library’s overall mission. In this column, I’ll look at some of the technologies and methods that facilitate cooperative resource sharing and describe some approaches in very broad and general terms. In the sidebar, I’ve listed Web links to more detailed information.
Libraries can increase the services they offer to their users through cooperative alliances. With few exceptions, no one library can collect all the literature in all areas of interest and academic disciplines. It makes good sense then for them to form relationships and collection-development strategies. In such an alliance, each library collects in its specialty areas and makes a conscious decision not to purchase extensive materials in other areas. But for this strategy to work, libraries must implement effective and efficient ways for their users to gain access to the materials they need, even if those materials reside in a partner library rather than the users’ local library.
Sharing a Common System
One of the basic technologies that supports resource sharing among library groups is an online union catalog that represents the collections of each of the individual libraries. In some cases, because all the libraries may employ the same library automation system, a single combined database might already exist. This is a typical library consortia model, through which users can easily search either the comprehensive union catalog or limit their queries to a single library. With this model, a library patron can use the automation system’s basic circulation features to request an item not available locally but held in a remote library within the consortium.
The shared-system, union-catalog approach works well for libraries that are closely knit, or that share a common funding agency. Union catalogs are quite common. School districts, state university systems, and even many statewide, multi-type library consortia rely on this approach. Libraries that have the luxury of participating in a network that shares a common automation system gain a great technical infrastructure for sharing materials, performing cooperative cataloging, and employing strategic collection development.
Connecting Diverse Systems
Libraries that are more loosely related may want to share resources as well. Such a library group will likely not have a shared database, and each library may use a different automation system. In these cases, a "virtual union catalog" can be created. All the major library automation systems on the market support the Z39.50 search-and-retrieval standard. With the right software, you-can use Z39.50 to create a union catalog among a set of diverse library automation systems. Such software needs to function as a Z39.50 client, have the ability to execute a search on multiple servers simultaneously, and collate the results. Using the Z39.50 protocol, the virtual catalog software submits the user’s search to each of the represented libraries’ Z39.50 catalogs and gathers the resulting records. In a basic system, the results from each library can be presented to the user separately. In a more sophisticated environment, all the results are combined, duplicated, and sorted.
One of the important features of a virtual catalog is the display of not just the bibliographic information in the results, but also the circulation status of each item with detailed copy, location, and holdings information. The ability to show the status information was a relatively late development within the Z39.50 standard. While many of the designers of Z39.50 applications had previously introduced various methods for displaying status and holdings information, it was only in the last couple of years that a standard holdings schema was officially incorporated into Z39.50. Among the current Z39.50 products, there are some that have already implemented this functionality using the now-standard holdings schema, while others follow approaches that were developed before the standard was established. In other words, the process of configuring a virtual catalog to correctly display circulation status and holdings information among a diverse group of library automation systems can be a complex process that invo lves a lot of tweaking and tuning. It’s important for the software that supports the virtual catalog to have the flexibility to deal with the various approaches used for delivering the status and holdings information.
Developing a consistent searching method among diverse library automation systems in a virtual catalog can also be complicated. Each of the library automation systems has its own characteristics related to search and retrieval. What specific fields are in a title, author, or subject index vary among library automation systems, for example. Given the high degree of customization in modern library automation systems, the ways in which each field in the MARC record is indexed can vary widely among libraries that use the same software. The different Z39.50 servers also vary in the search types they support. Some offer both title and phrase searching for each of the indexes, some offer only one or the other.
In response to the large degree of variability in searching within the Z39.50 framework, a tremendous amount of effort has been invested in designing a more consistent approach. The Bath Profile: An International Z39.50 Specification for Library Applications and Resource Discovery is the result of this initiative. A group of libraries has worked together to define in some detail a set of attributes that should be associated with each search type to ensure consistent results through Z39.50. If each of the developers of Z39.50 servers, and each library that implements a Z39.50 server, follow these specifications, it can finally become possible to implement a virtual catalog that produces consistent results from each of the participating systems.
Interlibrary Loan Issues
If the main interest in establishing a virtual catalog is to support resource sharing, then it’s important to offer library users the capability to efficiently obtain materials that aren’t available locally from the participating libraries. Thus the key feature of a virtual catalog environment is an integrated interlibrary loan (ILL) request capability. First of all, it should be clear from the virtual catalog whether the item that a user needs is held in his or her local library and if it’s currently available. It’s important to avoid the more time-consuming process of obtaining an item from a remote library if the materials are available locally. But when an item isn’t owned by the local library, or is currently checked out, an integrated ILL capability is valuable.
There are even more options and complications for ILL within a virtual catalog than those mentioned earlier with Z39.50. Should the system interoperate with OCLC’s interlibrary loan subsystem, or should it communicate in a peer-to-peer mode with the participating libraries through the ISO 101601/10161 international standard protocols? Will the ILL request be sent to the local library’s interlibrary loan office, or directly to one of the libraries that has been identified by the system as having the item available? Will each request be manually reviewed by ILL staff, or can the system be trusted to work automatically? How will materials be physically transported from the lending library to the requester? Do requests initiated through the virtual catalog take precedence over other ILL transactions? These are but a few of the issues that arise as libraries cooperate to share resources through a virtual catalog.
Virtual Catalog Projects
At Vanderbilt’s Heard Library, we’re involved with two different library alliances that are each supported by a virtual catalog. We participate in the Nashville Area Library Alliance (NALA), a group of about 18 libraries in and around Nashville, Tennessee. The group includes academic, public, school, and medical libraries. We also participate in IRIS, a partnership of the three major university libraries in our region.
Twelve of NALA’s 18 members are part of a virtual catalog called Athena. We use OCLC’s SiteSearch to manage the Athena catalog. SiteSearch is a very powerful and flexible tool, and includes all the features I mentioned earlier that are necessary to support a Web-based virtual catalog. This software is highly customizable, has a lot of flexibility in accommodating the variations in Z39.50 implementations, supports the presentation of circulation status and holdings information, and has integrated ILL support. The process of building and customizing applications with SiteSearch is done through the Java programming language.
The Athena catalog follows a centralized model with Vanderbilt serving as the single host site. None of the other participating libraries needs to have SiteSearch, they just direct their users to the central server’s URL. This centralized approach to the virtual catalog works well when the total volume of use remains fairly modest and when there isn’t a great need for major customizations by each of the libraries. Since the development and maintenance of the virtual catalog is supported by the host site, other libraries can easily take part even if they don’t have a large computing facility or abundant technical support staff.
General access to search the Athena catalog isn’t restricted, but a user must be authenticated as belonging to one of the participating institutions to be able to initiate an ILL request. Athena’s members are committed to providing expedited service for requests made through the Athena catalog. In general, an item can be obtained from another Athena library in just a day or two, rather than the normal, longer time frame for ILL requests. The group has engaged a courier service that makes daily rounds among the libraries to ensure prompt delivery of materials.
Vanderbilt also participates with the area’s two other large academic libraries--the University of Tennessee-Knoxville and the University of Kentucky-Lexington--in an another alliance called IRIS. Within this alliance the participants not only use a virtual catalog to share their existing resources, but are also in the process of reshaping their collection-development policies to minimize duplication among the three libraries.
The IRIS virtual catalog follows a distributed approach, rather than the centralized one we used for Athena. Each of the three IRIS libraries has its own implementation of the OCLC SiteSearch software, allowing each to customize its version of the catalog according to its local interests and concerns. Each can customize the catalog’s look and feel so that it’s consistent with its local online catalog, and each can deal with authentication and ILL in its own way. Each of the three institutions already owned SiteSearch at the time the project started, and each has adequate technical infrastructure and support staff.
As with the Athena group, the IRIS participants offer each other greatly expedited service for ILL requests. While the geographical area is significantly larger, the use of a courier service results in very short delivery times.
Through these two cooperative arrangements, we’ve greatly expanded the resources to which our users have access for prompt ILL service. We consider these efforts a great success and consistently receive positive comments from our users.
Marshall Breeding is the technology analyst at Vanderbilt University’s Heard Library and a writer and speaker on library technology issues.