The library community's intense interest in discovery interfaces that allow libraries to offer more modern tools to their users for searching their collections continues. While many libraries cannot make wholesale changes in their automation environment, many seek better interfaces for their end users. Almost all of the commercial library automation vendors now offer products in this genre, including Encore from Innovative Interfaces, AquaBrowser created by Medialab Solutions and distributed by R.R. Bowker, Primo from Ex Libris, LS2 PAC from The Library Corporation, Illuminar from AutoGraphics, BiblioCommons, and Enterprise from SirsiDynix. As fewer libraries purchase new ILS products, the sale of products like discovery interfaces represent an increasing proportion of revenue for these companies.
In addition to these products, which were created by commercial companies, a number of alternatives have emerged in the open source arena. These open source products allow libraries to follow a more experimental approach in establishing their nextgeneration discovery interface strategy. Libraries can download and install the software on a test server, load sample data sets, and customize it without the need to make a major commitment or financial outlay. They can even experiment with multiple products. The availability of these open source products allows a library to make an initial investigation of a new generation interface and gain hands-on experience, even if they might eventually decide to purchase a commercial product.
The number of libraries currently making commitments to open source discovery interfaces currently falls well below the number of libraries who have purchased commercial products. Yet, we still see momentum building toward open source versions of discovery products. The list of those going with open source discovery products now includes some very large and prestigious libraries. In the library automation arena, success builds on success. If these projects to implement open source discovery interfaces prove successful, they will pave the way for others. Just as in the ILS arena, commercial and open source alternatives will coexist as libraries seek products to replace their aging ILS OPACs.
This genre of discovery interfaces provides a new tool that library patrons can use for searching library collections. In most cases, the discovery interface operates somewhat separately from the ILS, though it interacts with it in many important ways. The ILS continues to serve as the automation environment for library staff as they manage the library's collection. The discovery products extract data from the ILS and other repositories that manage library collections, creating a new centralized index with new search capabilities.
The discovery interface often interacts in real time with the ILS in order to display information regarding the availability of materials and to place requests. In the open source discovery interface arena, VUfind stands as the dominant product. The Mellon-funded eXtensible Catalog project has been underway for about 2 years and will soon be releasing software. Other open source discovery technologies emerging include the University of Virginia's Blacklight and OpenBib, which is an example of discovery interfaces developed internationally.
VUfind was one of the first open source discovery interfaces for libraries, and it continues to be the dominant non-commercial alternative to proprietary products. SLN featured VuFind in its September 2007 edition while it was still in early stages of development. Since that time, the product has matured in functionality and has seen adoption both in its home institution and beyond. Version 1.0 was released around October 2008. In December 2008, the Andrew W. Mellon Foundation recognized Villanova University as one of the recipients of its third annual Mellon Award for Technology Collaboration, designating a $50,000 prize for the project.
A number of major libraries have implemented VuFind, with some now offering it as the default search interface. Others have created preliminary installations of VuFind, offering them to their users as an experimental alternative. As an open source application that sees adoption in a variety of major institutions, VuFind benefits from a growing community of developers.
Some of the libraries using VUfind now include:
The Falvey Memorial Library at Villanova University, which initiated the creation of VUfind and continues to lead its ongoing development. An early version of the software was created to provide access to a Community Bibliography (http://bibliography.library.villanova.edu/), and was later extended to address the library catalog, emerging as VUfind. The primary developers of VUfind included Andrew Nagy and Chris Barr. Villanova began using VUFind as the default interface for its Voyager ILS in August 2008. The ability for the other libraries listed below to make use of VUfind would not have been possible without the development carried out at Villanova.
The National University of Australia became the first major library to put VUfind into production as the default interface for its collection on May 27, 2008—even ahead of Villanova University. NLA uses VUfind in conjunction with its Voyager system, searching a collection of over 5 million titles. The library has recently expanded its scope to include all of the text of Project Gutenberg, a collection of books in the public domain, and the material in the Hathi Trust, a shared digital repository of over 2.6 million digitized volumes from 25 major universities in the United States. The use of VUfind by a library of the stature of the National Library of Australia gives the product significant credibility.
The Consortium of Academic and Research Libraries in Illinois, or CARLI, has implemented VUfind as an alternative interface for the I-Share catalog for its 76 member libraries. These libraries share a Voyager system, and offer both the native WebVoyage and VUfind interfaces. Some CARLI libraries already position it as the preferred interface. The adoption of VUfind by CARLI demonstrates its ability to address the needs of a large library consortium.
The Minnesota State Colleges and Universities system operates a shared Aleph implementation for its member libraries, called MnPALS. The organization has implemented VuFind as an alternative interface under the name MnPALS Plus. Considered a production product, some PALS members currently position it as their primary interface, with additional libraries planning the transition in the next year.
Western Michigan University plans to offer VuFind as its primary search interface by Fall 2009, replacing its current Voyager catalog. VUfind was selected following a process that evaluated Primo, Encore, Endeca, AquaBrowser, and WorldCat Local.
Several members of the Association of Research Libraries have shown interest in VUfind. ARL members present significant demands for automation software given their large collections and complex organizations. Georgia Tech University, for example, relies on it as the default interface for its Voyager system. Colorado State University launched its Discovery search tool in February 2009, based on VUfind. At CSU, VUfind operates with a Millennium system from Innovative Interfaces. Auburn University implemented a test version that it expects to put forward for user testing in Summer 2009. The University of Michigan, an Ex Libris Aleph site, offers an alternative version of its Mirlyn catalog based on VUfind. Yale University has implementation a version of its catalog based on VUfind it calls Yufind. Yale has done extensive work testing the usability of its implementation of VUfind.
York University announced that it selected VUfind as the preferred discovery interface. At York, an evaluation team was charged with evaluating the discovery products currently available, and ultimately recommended VUfind as its recommendation. Implementation is planned for Summer 2009. The total number of libraries involved with VUfind or other open source discovery interfaces represents a small portion of the 123 total members. Yet, the ability to break into the ranks of this group at all represents a major milestone for an automation product in the academic library sector.
In Germany, the Verbundzentrale de Gemeinsamer Bibliotheksverbund has used VUfind as the basis of a search and discovery interface for a large collection of scientific and technical resources called Nationallizenzen. The current beta version indexes 4.6 million items, but will grow to a total of over 50 million. This resource includes about 1 million items freely available as well as proprietary resources that require authentication. This resource is funded by DFG, the German Research Foundation, one of the major sources of funds for scientific research in Germany. This implementation of VUfind illustrates its ability to address collections of content other than traditional library catalogs.
In other news related to VUfind, Andrew Nagy, its primary developer at the Falvey Memorial Library at Villanova University has recently joined Serials Solutions as Senior Discovery Services Engineer, focusing in that company's new Summon discovery product. Nagy will continue his role as the lead developer of VUfind.
At this stage, VUfind ranks as an established competitor in the discovery interface product genre. While not as many libraries use it as their production interface as some of the commercial products, it is off to a respectable start. The projects listed here do not represent a comprehensive list of libraries working with VUfind, but a selection of examples that illustrate its use in diverse library settings.
eXtensible Catalog reaches milestones
The eXtensible Catalog (or XC) project continues its progress in creating an open source discovery platform for libraries. SLN covered this project in its December 2007 issue, following its award of a second round of funding from the Andrew W. Mellon Foundation. Since that time, progress on the project has taken place mostly behind the scenes. Although the XC team has been active in describing the conceptual framework behind its approach, no software has been released with specific information about what the eXtensible Catalog will look like.
One of the key concepts behind XC involves its approach to metadata. XC includes a Metadata Toolkit for exporting and transforming metadata into forms that make it more effective in a faceted search environment. It also includes toolkits for OAI and NCIP for interacting with ILS systems and other repositories. XC will embrace multiple interface options, including one that uses Drupal, an open source content management system that finds widespread use, with an increasing presence among libraries. By working with Drupal, XC will be able to take advantage of a mature set of interface tools, which can be extended to incorporate library data and services. Alternative interfaces for XC will be created for Learning Management Systems so that academic libraries can more easily present their services through these environments.
The XC project will soon begin making its software available. The University of Rochester River Campus Libraries has recently announced that it plans to release early versions of the various components that comprise XC by March 30, 2008. The release of the software and documentation does not signal completion of the project. Development will continue, but in a more public way that will allow the library community opportunities for review and comments.
Interest in open source development of new library interfaces extends internationally. The University of Cologne in Germany, for example, has uses a locally-developed search environment called Koelner UniversitaetsGesamtkatalog, or KUG. This portal provides access to the holdings of the 145 institutes associated with the University, each of which maintains its own catalog, as well as other repositories and special collections. KUG currently searches about 7.1 million items. This project relies on OpenBib, open source software for library interfaces whose original development began in 1997 led by Oliver Flimm. It makes use of open source components including Apache, Perl, and MySQL. OpenBib uses the open source Xapian search engine toolkit (xapian.org), unlike many of the other library discovery products that tend to use Lucene.
OpenBib includes the standard features of the current line of library discovery interfaces, including faceted searching, end-user tagging, reviews, tag clouds, and recommendations of popular or related items based on use data. The interface offers a live-search feature, where the system begins to offer search terms in a clickable dropdown that builds as the user types a few characters. This increasingly popular feature not only saves time in typing, but presents the user with valid search terms that might not have otherwise been obvious. OpenBib includes the ability to browse the collections, allowing the user to drill down from broad disciplines to structured subject terms down to specific works. OpenBib makes generous use of RSS, offering feeds for subject terms and authors represented in the collection, as well as search results. A new version of KUG has been developed that looks more similar to the current line of discovery interfaces. This new interface, now available in a beta test version, will become the default interface in the near future.
So far, no other libraries are using OpenBib to power their catalogs other than the University of Cologne. For more information OpenBib, See: http://www.openbib.org. KUG : http://kug.ub.uni-koeln.de/
Another open source discovery product that has begun to receive attention is Blacklight, a development project at the University of Virginia (UVa) libraries. Blacklight was created as a research project by the University of Virginia and has not yet been put into production for the catalog of the UVa Library or any other library. Like VUfind, Blacklight is an open source faceted discovery tool based on the Apache SOLR technology. It relies on Ruby on Rails as the programming language for presenting its user interface. Blacklight emerged out of a project to create a tool called Collex that it originally created to provide access to a collection called NINES, nineteenth-century studies online. Blacklight operates with a variety of international metadata formats including MARC, EAD, TEI, and General Descriptive Modeling Scheme (GDMS) developed at UVa. Blacklight offers a very similar approach to VUfind, with comparable features. The Ruby on Rails programming environment has become very popular for Web development. While still a research project, Blacklight serves as an example of ongoing work in the discovery interfaces arena that will lead to even more of a variety of options and opportunities in this arena in the future.