The Advent of Portals

The internet, digital electronic resources, and database technology have transformed the way people search for information.

Today many people rely on keyword searches in single-step search engines--e.g., Google or AltaVista--that retrieve information from unspecified slices of the web. This is eroding the use of traditional library reference and information services and could ultimately result in a set of services and resources that are less authoritative but more convenient.

Users frequently cite ease and convenience as the main reasons they prefer commercial search engines over gaining access to electronic sources through a library's web site. Sources such as online journals, online public domain materials, or locally developed databases are often passed by because of these convenient options. Libraries must gear up to provide a competing level of convenience while retaining the authority and quality of information delivery for which they have been traditionally known.

For a host of queries, of course, the Google or AltaVista search may be all a user needs. Librarians must reconcile themselves to this fact and refocus the mission of library information services and resources to the deeper, more complex information needs of users engaged in searches that require higher levels of authority and more comprehensive scope.

Imagine one web site that can combine the powerful searching of web resources with the searching of local catalogs, online journals, or locally digitized resources. Add to this the ability to initiate a reference question, submit an interlibrary loan (ILL) request, and transfer into course management systems a citation or portion of a journal article, all without leaving that web site.

While the library and IT communities have not agreed on a single definition of a portal, there is growing consensus on the essential features and functions of one.

Michael Looney, cofounder of the portal company GoCampus, Inc., and Peter Lyman, at UC-Berkeley's School of Information Management and Systems, offer one definition: 'Portals gather a variety of useful information resources into a single, 'one-stop' web page...[that] allow users to customize their information sources by selecting and viewing information they find personally useful.'

Like Looney and Lyman, Andrew Cox and Robin Yeates, both associate directors of the UK's LITC, expect portals 'to integrate the diverse licensed and owned electronic holdings of libraries for users, through the whole process of discovery and searching to final delivery, regardless of the content's format, the metadata standard in use, publisher interface, or authentication mechanism.'

Sarah Michalak, director of the University of Utah Libraries, has defined a 'dream portal' as a super discovery tool that specializes in high-quality content. The dream portal is fast and powerful. It searches across formats and resources and returns results that are deduped and relevancy ranked. It is more than a discovery tool because it delivers full text or information objects whenever available. The dream portal integrates appropriate applications such as course management software. Finally, the dream portal supports authentication and permits customization and personalization, e.g., alerts, saved hits or searches, and custom views of resources. This dream portal will help users overcome 'infoglut.' It is 'Google with good content supported by a range of library services.' A portal combines powerful searching with the diverse resources and services that patrons find when they use a library. Portals should provide library experience of that quality without requiring people to come to the library.

Portals are more than enhanced web pages, although some have evolved from library web sites. Librarians are just beginning to define the requirements for portal products. The portals and librarian expectations of their functionality will become more refined and sophisticated as libraries adopt them. Users may 'discover' the library portal from a Google search, a visit to the library, or in any of the ways users now discover new resources in libraries. The core feature of any portal will be integrated, cross-database searching of a local catalog, other library catalogs, selected web sites, locally licensed full-text and abstracting/indexing databases, and public domain or publicly accessible abstracting and indexing services.

Integrated searching is a key feature of a portal. It distinguishes it from a web site. Many library web sites provide access to the online catalog, licensed resources, vetted web sites, and links to one or more commercial search engines. However, access to these disparate resources is most frequently accomplished by searching one source at a time. Library web sites usually do not permit users to conduct a single search of multiple resources, nor do web sites deliver integrated results. Users usually have to integrate the results from their separate searches as another step.

Both the multiplicity of standards and the lack of standards are challenges in developing integrated, cross-database searching. Many online catalogs can be accessed by the international standard for search and retrieval, Z39.50, but additional search techniques are required for such resources as XML datasets or web resources using different metadata schemes such as MARC, Dublin Core, Computer Interchange of Museum Information (CIMI), and Encoded Archival Description (EAD).

Keyword searching is common for web sites. Licensed resources may have proprietary search strategies. As a result, portals must support various search standards and protocols (Z39.50 and http), and they must integrate the results. Portals also must support a variety of controlled vocabulary or thesauri. Library users accustomed to searching Medline, for example, will expect comparable results if they search Medline through a portal interface.

Commercial search engines like Google present results ranked by relevance. Users, however, often prefer one kind of relevancy over another. One user may want full-text resources displayed first. A second may opt for results from a specific journal or resource. A third user might choose all materials held locally. Portals must be able to rank search results differently to meet the needs of different users. Users should be given choices in how those results are ranked or listed. Some users, for example, may want only items from one source while others will want to see only a listing by date of publication.

While discovering an information citation or resource is obviously a core function of a portal, it is not enough for portals to be just search engines. Users want to use the information they discover and that means portals must provide for that use. The portal's supporting services must supply the ability to capture, integrate, manipulate, and distribute the information and offer ways to consult others and collaborate with them in the process.

At the minimum, users will expect to capture the information resource and bring it to their desktops. Most people prefer to get the full-text or full-image object rather than simply a citation. If the electronic document is not available, then capturing the citation is necessary. Portals will increasingly offer the functionality to read an OpenURL, transfer bibliographic or descriptive metadata, and check that the user has suitable permissions to access each relevant resource and then enable links to applicable resources. The OpenURL standard also enables dynamic linking to local copies of electronic journal articles, library catalog records, and remote commercial article services. It also helps maintain static URLs, seamlessly directing users to the most appropriate copy.

Another new open standard will permit users to access their circulation or ILL records from the portal interface. NISO members approved the NISO Circulation Interchange Protocol (NCIP) standard in July 2002. When NCIP is incorporated into portals, users will be able to place holds, recall items, or check on the status of ILL requests.

Portals can also offer the ability to transfer captured citations into ILL requests, commercial document delivery requests, or requests for the library to purchase the item. Portals that do this eliminate the current need to search in one tool and enter ILL requests using another. It is the seamless flow from discovery to requesting that will make for a successful portal.

After capturing the information, users expect to be able to integrate some or all of it into a variety of related applications. For example, users may wish to put images from the Library of Congress's American Memory Project into class assignments created within Blackboard or WebCT, perhaps including an audio clip relating to the image. Users expect to manipulate the captured information objects or citations by excerpting text, annotating citations, creating bibliographies, and manipulating images. Integration with local e-mail systems, calendars, and campus schedules must be part of new portals.

Students in many universities now expect to learn in a collaborative environment, either physical ones such as the Information Commons at the University of Arizona Libraries or Emory University Libraries or electronic ones such as chat rooms or shared workspaces. In a portal environment, students will expect to share information resources with their classmates, and faculty will expect to distribute class assignments, engage in chat-room discussions, and provide pointers to information resources. Portals that seamlessly link to learning management environments will bring library-vetted resources to students and faculty rather than forcing them to use the library web site to search for materials.

Links to virtual reference services will permit users to seek the help of reference librarians when they need it and without physically going to the library. Portals will be viewed as virtual libraries, with a range of services equal to those currently provided in the library.

Library portals cannot be developed in isolation. They must be interactive with many other systems, including university portals, content and course management systems, and document management systems. For academic institutions, the library portal must link to the learning management systems (LMS) or course management software. So far there has been little interaction between these systems. Learning management systems such as Blackboard and WebCT have been developed with minimal input from librarians. These systems support a variety of functions, including access to faculty-developed course materials, required and optional readings, chat rooms, and grades.

One area of potential integration is for LMSs to access electronic content licensed by the library rather than having the faculty or department license identical resources from a commercial supplier. Transfer of data between a portal and an LMS should result in seamless interaction between the two applications. For example, when students are searching the library portal, they should be able to transfer citations as well as electronic resources from the online catalog or other resources directly into the LMS. Likewise, students should be able to search the online catalog and other library resources from within the LMS and find references to resources or the electronic resources themselves.

Many academic institutions are developing university or college portals. Some view the library portal as simply a channel on the university portal, while others view the library portal as parallel to the university portal. Both alternatives will evolve, and regardless of which option predominates, access to information resources will be available to information seekers.

Research libraries, like academic, public, and special libraries, view library portals as an enabling tool to support the mission of their institutions. Librarians have high expectations for how portals will facilitate access to a wide range of high-quality content. The number of companies marketing portal products is expanding rapidly, and the number of libraries offering portals with integrated searching and multiple supporting services will continue to grow. The challenge and the fun ahead is to design portals with enough flexibility to respond continually to user preferences in the discovery, presentation, and use of high-quality information resources.