Library Portals and What's Next

• By John Savarese
• 05/05/05

Is the library portal about to yield to the Next Big Thing?

A funny thing happened to the traditional library card catalog as it morphed into the online public access catalog (OPAC). The whole mission of the library changed along the way, and the purpose of the library catalog changed along with it. Enter the library portal.

But, first: How did we get here? The file-drawer-based card catalog was an inventory of the objects (mostly books and periodicals) that were physically present in the library. With the burst of new electronic resources-first electronic databases on CD, then eJournals (full-text electronic collections), and all manner of resources on the Internet-the library's job got a lot bigger. People who used the library expected it to help them deal with the entire universe of information, not just the books on the nearby shelves. So, the OPAC had to transform itself to meet a similar challenge.

VIA MY LIBRARY, LEHIGH STUDENTS
can now build their own
customized resource portals
with material from the
Linderman Library (left)
--or anywhere else.

The online library software somehow had to catalog, or at least provide a gateway to, a huge variety of materials in the hands of many different proprietors. Integrated Library Systems (ILSs) have met this challenge by converting the simple OPAC into a comprehensive, customer-driven library portal. (See the box below for a sampler of portal-style features offered by various vendors and institutions.)

Here, in a nutshell, is what makes a library portal different from the OPAC of old:

Federated search. A well-stocked library may subscribe to hundreds of online databases and other resources on behalf of its patrons, each with its own search interface and login procedure. A federated search lets the user enter the search criteria once and eliminates duplication among the results.

User profiles and contexts. The system knows who the user is and what the user generally wants, and uses that information to tailor its services, integrating with the campus's administrative information and course management systems.

Multiple channels of content. The system can offer weather reports, RSS feeds, and the dining hall menu, in addition to more formal library databases and collections.

Customizable content and interface. The library can customize the portal by branding it with its own look and feel. Users can choose interface design and needed tools, by default. Some library portals even offer the flexibility of "skins": easily interchangeable surface designs such as those featured by MP3 software.

The Crystal Ball

Stephen Abram, VP of Innovation for the Sirsi Corporation (www.sirsi.com), is a seer in the library automation field. (See his articles at www.sirsi.com/Resources/abram_articles.html .) He says that Sirsi, for one, is building a "framework" for college and university library sector success. For instance, the company will be introducing a product this year called Enterprise Portal Solution (EPS), built around the concept of "rooms." A Sirsi Room, says Abram, is a context that a user is working in at the moment, such as cell biology or finance. With EPS, the university library will be able to assemble the right resources in that room to prevent information overload (hundreds of resources that may not be relevant), and increase the likelihood that the user will find solid, reliable information. Other areas the company is working on include: closer integration with course management systems like Blackboard (www.blackboard.com) and WebCT (www.webct.com); an instant-messaging-style virtual reference desk; profiles linked to a user's library card that automatically invoke accessibility software matching the user's special needs; federated identity management so that users only need to log into the campus systems once; and support for experience-based learning styles rather than just text-based ones (accessible today through Sirsi's Hyperion streaming media server; in the future, perhaps, nurses will be able to interact with a virtual cadaver).

Susan Wagner, Sirsi's Northeast account manager, claims the company is putting "content in context." She says that Sirsi is merely "bringing to the table what librarians have brought to the table for hundreds of years: the ability to distinguish between relevant material and junk"-only now in a virtual environment.

U.S. PORTAL AND FEDERATED SEARCH FEATURES

Creating Context

Lehigh University (PA) is creating context for its library users with the help of an open source library portal called MyLibrary that is being developed at University Notre Dame (IN) under the leadership of Eric Lease Morgan, head of Digital Access and Information Architecture. At Lehigh, MyLibrary collects information about each user (such as a major declared in the university's administrative information system), and uses it to build a customized collection of relevant reference resources, which the user can then customize to taste. MyLibrary also interacts with the campus's ILS to automatically retrieve references to journals and databases in the user's field of interest.

Faculty, graduate students, and undergraduates at Lehigh have already used MyLibrary more than its planners ever projected. Tim McGeary, senior library systems specialist, explains Lehigh's success with MyLibrary, which is integrated with the university's SunGard SCT Luminis campus portal (www.sungardsct.com): "Before students ever step on the campus, they do all their summer orientation through the Luminis portal. Their Lehigh experience is born through the portal. If they accept that interface as being the way to go, then a portal-based service like MyLibrary will succeed."

Integration, Interfaces, and Irony

D'es a comprehensive library portal run the danger of overlapping and competing with efforts to develop an effective institutional portal? Doug Randall, technology product manager at Innovative Interfaces Inc. (the developers of the Millenium ILS; www.iii.com), sees the two kinds of efforts eventually converging. "Institutional portals can seem sometimes to subsume the library, by implementing some of the federated searching and metasearching that the library may already be offering to its users. Where you draw the line is not clear. But the future is probably integration rather than fundamental migration of the function- ality totally in one direction or the other."

In fact, much of Innovative's recent efforts have been aimed at making integration between Millenium's portal and other campus systems easier. Efforts include support for single sign-on technology based on Apache authentication modules like Central Authentication Service (CAS) and Pubcookie; the development of portlets; a project to implement RSS feeds within Millenium; and the development of a mini-search widget (a small, free-standing application that will operate outside the Millenium interface).

But what about the future of the portal concept itself? There may be a huge irony here. The primary concept that makes the portal possible is that information services can be constructed in a way that makes them independent from a specific appearance they must have on screen. There are many names and styles for this abstraction: Web Services/XML, OpenURL, APIs, widgets, and RSS feeds. These "browserless" services provide information when a program asks for it; a user d'esn't need to point and click. The information is returned in a structured form that can be easily processed by the local program that called for it.

Once services like specialized search engines are provided this way, the portal can freely combine and repackage them. But now the horse is out of the barn, because the user can begin to choose from many interfaces to reach these abstracted services. The interface then becomes a commodity, not a monopoly. As a user, as long as what I want is available in the form of a service, I can build my own portal (or buy one off the shelf) using my own desktop rather than someone else's Web server, whether it is the library's ILS portal or the institution's official one. Even a federated search backed by strict authentication can be offered as a service, if the provider wishes to do so.

You may already have installed software on your computer that links you up to the recent crop of consumer- oriented browserless services, programs as simple as MediaMan (www.imediaman.com), Konfabulator (www.konfabulator.com), or an RSS feed reader. MediaMan catalogs your personal collection of books, recordings, and other media by harvesting information (including book jackets) from the browserless Amazon Web Service (www.amazon.com). Other Web services are already being made available for software developers to use by eBay (developer.ebay.com/DevProgram/index.asp) and Gracenote CDDB music information service (www.gracenote.com).

Users will still desire a place where they can assemble their favorite services. But in the world of browserless services there will be multiple, side-by-side solutions, some of them on the client side rather than on the server side. Certainly, institutional and library portals will continue to play an essential role, but the portals that users value in the long run may be the ones that they own and operate themselves.