L L E T I N
Portal Implementation Issues and Challenges
Eric Lease Morgan is head of digital access and information architecture, University Libraries of Notre Dame. He can be reached at firstname.lastname@example.org.
If you think librarianship is about the collection, organization, preservation and dissemination of information and knowledge, then implementing a library portal will be easy. On the other hand, if you think librarianship is about books, subscriptions, MARC records and AACR2, or integrated library systems – the physical manifestations of traditional libraries – then you will have a hard time. This essay outlines some of the challenges of creating a library portal. The keys to overcoming these challenges, in brief, are not technological but philosophical and interpersonal. When there is a clear, shared vision of what the portal is intended to accomplish the rest falls into place.
Working Inside a Library
Communication, learning curves and new work, quality control, planning for the unexpected, the value of supportive partners – we must consider all these things inside any work environment, including libraries. Communication is one of the most important of these factors. Get out of your office. Walk the halls. Chat with your neighbors. Avoid email. Employ the telephone. Nothing happens without communication.
The information/knowledge environment is drastically different with the advent of the Internet. The changes it has brought have altered people's expectations, and libraries are increasingly becoming irrelevant because they are too slow to adapt to the environment. We must do things differently, which requires accepting new work and climbing steep learning curves. Create databases – don't just search them. Adapt to everybody else's data and communication structures (XML and HTTP) – don't needlessly cling to your own (MARC and Z39.50). Learn to syndicate your content – don't force people come to your physical space. Although planning for the unexpected is a seemingly impossible task, being adaptable to change makes such things much easier. Surprises become expected.
Quality control is important in any environment. As in the phrase controlled vocabulary, the operative word is not "control" but "quality." Selectivity is important, especially in these days when access to information is almost trivial. Articulate the characteristics of the information and knowledge you want to include in your library (read "portal"). This articulation becomes your definition of quality. Use that definition to selectively create your collections and implement your services. Prioritize your selections, making it easy for users to choose. As Mr. R. used to say, "Save the time of the reader."
Supportive partners are the result of building interpersonal relationships. These relationships are not only between you and the people in your immediate physical area, but people outside your department, your division, your library and even your institution. "No man is an island," and what you do affects everybody else. The creation of these relationships begins with communication, continues with shared goals and matures through cooperation. When building your library portal, keep these things in mind. Sincerely try to understand the other person's point of view.
User-centered design is an underlying principle for any library application, including portals. A sort of Copernican Revolution is becoming apparent in libraries. Now more than ever, libraries are no longer the center of the information universe. The focus has shifted to the user as the driving force behind library services, not the Platonic ideals of information organization. User-centered design is the manifestation of this shift. If your services are not user-centered, then your services will not be used.
The key to creating successful, user-centered services is providing services users want and desire, not necessarily services librarians think they need. To learn what users want and desire conduct surveys, focus group interviews and usability studies. Ask and listen. Whether we believe it or not, users are much more information literate than we suspect. Make sincere efforts to implement services to meet users' desires. Ask yourself what tasks your users want to accomplish. Do they want to find a list of articles on a particular topic; read a particular journal article; renew a borrowed book? Implement those services. Make them easy to use without requiring bibliographic instruction sessions. Facilitate usability tests. If users can accomplish the tasks, then great. If not, then it is back to the drawing board.
Libraries are always a part of a hosting institution. School libraries are a part of schools. Public libraries are a part of municipalities. Academic libraries are a part of colleges and universities. Special libraries are a part of companies, hospitals or governments. These hosting institutions fund libraries, and decision makers ask themselves "Have I used the library recently? What has it done for me?" The more difficult it is for the decision-makers and their advisors to answer these questions, the more difficult it will be for libraries to acquire funding. Implement the collections and services users want, and their questions will be easy to answer. Practice user-centered design.
Identity management is an essential part of a portal implementation especially since it is one of the defining characteristics of the application. In some ways, portal applications and digital library services in general need to integrate aspects of the traditional library reference interview into their implementation. The system needs to know a bit about the users before useful services can be provided. Have the people used the system before, and if so, what have they done and what did they find useful? What sort of information do the users desire, and therefore what resources might be suggested before others? What sort of expertise or previous knowledge do the users bring to the information need? The answers to these questions go beyond the typical administrative computer questions of authentication and access control, and the process of answering these questions in an automated environment is a challenge for the profession.
By comparison, the issues of authorization and access control are simple, but the key to implementation is integration with your library's hosting institution. Users have many important characteristics that go beyond name, address, telephone number and what rank they hold in the organization. Much of this information can be gleaned from the wider environment. In an academic setting, for instance, data such as what classes the person teaches or is taking may be apropos. The sorts of things the person has written can be valuable information for efficiently satisfying information needs. With the advent of institution-wide repositories, this sort of information is increasingly accessible. Real identity management goes beyond privacy.
Privacy is more a legal and self-imposed philosophical issue than a technical one. The difficulty is easily resolved when users understand that librarians take privacy as seriously as doctors and lawyers do and when that understanding is coupled with electronic waivers notifying users that information about them needs to be garnered and saved in order to provide the sorts of services they expect.
The question always comes up, "Should we implement an open source solution, build our portal by ourselves or license an off-the-shelf product?" The answer is determined by comparing the quantity of resources you have (time, money and people) with your tolerance for the unknown when it comes to computing. On one hand, if you have infinite resources, there is no reason why you shouldn't build your own solution. After all every institution is different, and one size does not fit all. On the other hand, if you are perfectly satisfied with your selected commercial vendor and their product does absolutely everything you need with no problems whatsoever, then an off-the-shelf solution is the way to go. The real world lies somewhere in between these two extremes. No one has infinite resources, and no vendor provides the perfect solution.
Even if you don't have infinite resources, there are good reasons to consider open source software or building something on your own. First and foremost, the development of open source software improves your ability to be computer platform independent. It is more likely to be standards compliant, and its development not only provides means for staff development, but also improves the profession as a whole.
Commercial software has the advantage of formal, institutional support. There is usually someone you can call. The implementation of commercial software requires less computing expertise by the library or the hosting institution, but if you ask me this is a false argument. The implementation of commercial software may allow you to focus on your tasks at hand and outsource the things that are not your core business.
The collection, organization, preservation and dissemination of information and knowledge are the goals of librarianship. They are never-ending processes requiring changes in implementation as environments change. Books, MARC and libraries defined as physical spaces make less sense today when compared to the environment 20 years ago. Portals, whether they are akin to metasearch initiatives like the Scholars Portal or the more personalized/customizable interfaces such as MyLibrary, represent new ways of providing traditional service and access to collections. Fostering relationships, practicing user-centered design, managing user identities and implementing computer technology are the sorts of things librarians need to consider when practicing librarianship today.
Copyright © 2004, American Society for Information Science and Technology