Eric Lease Morgan advocates the creation and maintenance of personalized, customizable interfaces to sets of library information resources such as the ones being developed over the last eighteen months. To accomplish this goal, this article outlines the goals of such systems, suggests a number of technical infrastructures for support, and describes some of the more important content and staffing issues.
In the Fall of 1997 the NCSU Libraries, in Raleigh, North Carolina, conducted a set of focus group interviews with students and faculty. These interviews were intended to help the Libraries discover how to provide better "digital library" services to its clientele. Three main conclusions were drawn from this study. First, half of our population thought they did not need a librarian because they believed a computer could fulfill such a role. On the other hand, the other half of the population thought they did need a librarian because they required a person to whom to discuss ideas about information. Second, half of our population thought they did not need a physical library, again, because they believed a computer could fulfill this function. And again, the other half thought there was a need for a library, a place, because they required a space to share and discover information and knowledge. Third, everybody stated they needed total access to the total universe of information, but, they emphatically stated, they only want to see one particular part of that universe at any one particular time. In short, our clientele were suffering from information overload. A similar study done at Cornell University, in Ithaca, New York had similar findings as alluded to by Angela K. Horne [1]:
Cornell's MyLibrary project began partially in response to a focus group study of Library Gateway use. It became clear that certain patrons were becoming overwhelmed by information excess, an understandable reaction given that we have 19 libraries and thousands of resources. The Executive Summary of the final report noted that users desired "a more personal space-- where they can dictate how important certain resources are based on their own needs." This was a motivational finding for us.
At the same time, services such as My Excite, My Yahoo, My Deja News, and My Netscape were appearing on the scene. These services allowed users to select and deselect content to appear on personal pages. The content usually included wire service news, weather, stocks, horoscopes, and links to Internet resources, along with advertisements. The information was useful, but all in all, rather generic and not necessarily scholarly. We live and work in an environment of increasing expectations, expectations concerning access, availability, and specialization. It is an environment driven by consumerism. When you go to the grocery store how many different types of beer are available? How about potato chips? You've got ruffles and no ruffles. There are sour cream and plain chips. There are sour cream with ruffles and sour cream without ruffles. There are sour cream with ruffles in tube, and sour cream with ruffles in bag. The choices go on and on and on, what's more, we expect these choices. It is in such an environment, an environment where information is so highly valued, plentiful, and expectations about availability are on the rise, where the personalized, customized interfaces to library resources develop. It is an environment of designer information and where information is perceived as a commodity.
Personalized library interfaces, "My Library" systems as they are often called, are an attempt to meet the expectations of the current information environment in a library setting. The goals of any My Library are essentially the same goals of traditional library service, to collect, evaluate, organize, and disseminate sets of information for the purposes of satisfying the information needs of library patrons. The essential difference is the medium in which this is done. In the present context, the sets of information are presented through a Web browser. Through a Web-based personalized interface libraries want to present to the user access to the totality of available information, but only relevant sets of that information at any one time. As Mr. R would say, saving the time of the reader.
Reducing information overload is seen as one of the most important goals of providing personalized library interfaces. For example, Norm Medeiros [2] assisted in the development of a MySQL-based program "To help our users circumvent wading through ever-growing lists of electronic resources." Amos Lakos [3] said "a well designed personal interface keeps the main contact local but enables access to the whole - but mainly it helps the student better and more useful use of time." Dan Ream [4] says such a system "allows me to quickly create a class-specific web page that my students can easily login to later when they're trying to remember where those databases were that the librarian showed them before." Again, saving the time of the reader.
Because information only becomes useful within some sort of context, the interface needs to provide that context or there needs to people readily available to help provide the context. In other words, the interface needs to provide access to people as well as information. Context is a necessary ingredient in an information world. It is the catalyst making information useful. The number 1776 has no meaning unless it set in some sort of context. Am I referring to types of numbers (integers, real, positive, even, etc.)? Does the number refer to the quantity of something such as the number of eggs I ate in the ten past years? If the number refers to a date, then suddenly it might take on a whole new meaning. Information must be set in context before it has value. Organizing information creates a context. Organization creates patterns, patterns provide the means for predictability, and predictability leads to the creation of new knowledge. Because librarians specialize in organizing large amounts disparate information into cognitive wholes, librarians can assist in the creation of context by placing similar information together. These cognitive wholes or "information cosmologies" as I like to call them are never perfect, but they serve to stimulate people's thinking about particular pieces of information and allow people to create and refine their own context. Ideally, any My Library system should help to facilitate the creation of this context too [5].
My Library systems must satisfy the needs of different people at different times as well as satisfy the needs of the same people but purposes. For example, one person may not need access to certain types of information such as bibliographic databases or statistical data sets while this sort of information is very important to others. Therefore, the system should accommodate these different people. Similarly, people should be able to rearrange the information on their page to meet their particular needs and computing styles. An instructor might use a My Library interface to regularly locate information on a particular topic, but that same instructor might want to create and "publish" a My Library page for their students to use in various classes.
As described below, every My Library system is essentially a database application with a Web front-end. Used effectively, this means an institution can describe an information resource in one place and have it appear in many places via different reports. Consequently, maintaining many Web pages is reduced to updating single records in databases and running the reports. As Anne Gambles [6] says:
Many academic librarians are currently spending much time maintaining resource information on typically "flat" library web pages. ... The RDM is a relational database which has enough complexity to cope with multiple instances of the same resource (eg in the case of a journal: hard copy, full text via an online database, abstracts via a stand-alone cd-rom) each with different levels of user access.
Norm Medeiros [2] is feeling better because of the database approach:
Prior to creating the database, approximately 2,500 e-resources were maintained manually in title- and subject-specific files. Every time we added, deleted, or updated a link, at least 3 HTML pages were edited. The database lets us make these changes easily and create new pages automatically. My fingers are grateful.
All of this begs the question about our online catalogs. Aren't they databases? Why can't we create My Library systems from that content? Part of the problem may be our catalogs' definition. Are they primarily inventory lists or finding aids? In the words of many a patron, "What is the OPAC?"
My Library systems can create a brand for individual libraries. The idea of a personalized, customized view of information is not unique to libraries. My Yahoo, My Excite, and My Netscape are well known examples. In fact, you can hardly throw a stick on the Internet without hitting yet another "MyThis" or a "MyThat", terms coined by Horne. Thus, there is a lot of competition for attention between customizable portal applications, and while most people think a customized interface to a library's collection of information resources is a good thing, a few people wonder how customized library interfaces stand in relation to other interfaces such as company intranets, university interfaces, and/or "portal-in-a-box" applications sold commercially. Debra Ketchell [7] said, "I like myLibrary and we are considering implementation as I have strong feelings about branding the library, but I am concerned about creating separate islands." Horne replies, "We aren't trying to push daily horoscopes at our students and faculty, rather develop a research webspace that is theirs alone and into which they can choose to add astrological information." Keith Morgan [8] echoes these sentiments, "I too want to brand the library through both a well designed public web site and options like MyLibrary but wonder if we will be overtaken by corporate forces. Or is the library environment so specialized that this would never happen?"
To summarize, the goals of a My Library system is to:
The balance of the article outlines the process of creating and maintaining a My Library system.
There are quite a number of technical infrastructures allowing for the creation of a My Library system, and they all rely on some sort of underlying database application. This is because the system must "remember" the user and for a computer to do this information about the user must be saved for future reference. Put another way, all personalized library interfaces are essentially database applications with Web front-ends. To illustrate this concept, consider the block diagram.
 Figure 1: Conceptual infrastructure of a My Library system |
The technical foundation of any system will be some sort of hardware, the host computer itself. The computer needs to run an operating system, the second level. A database application is installed for the purposes of "remembering" patrons. A programming/scripting language is the next level of the system and it is used to read and write to the database as well as provide some context for the system. The HTTP server, Internet connection, HTTP client trio provide the user interface to the programming/scripting language. Finally, people, either patrons or librarians, interact with the HTTP client to use and manage the entire system, respectively.
If a hypothetical system were built according to open standards, then each of the functional blocks in the system should be interchangeable with new technologies performing the same functionality as they became available. For example, if you were using Apache as your HTTP server, you should be able to substitute Netscape's Enterprise HTTP server if it served your needs better. Similarly, as long as your database supports standard SQL and/or ODBC, then it shouldn't matter whether or not you are using Sybase, Oracle, or MySQL as your database application. The most defining block of the system is the application/scripting language. Since this aspect of the system provides much of the context, the application/scripting language manifests the system's "personality". It is also within this block were you find the most diversity among existing My Library systems.
For example, the production version of MyLibrary@NCState runs on hardware built by Sun Microsystems (SPARC), the Solaris operating system, and Netscape's Enterprise HTTP server. On the other hand, the development version of MyLibrary@NCState runs on an Intel processor, the Linux operating system, and uses Apache as the HTTP server. In both cases, MyLibrary@NCState is implemented with the same set of Perl scripts. It is these scripts that give MyLibrary@NCState is look and feel, as well as it basic functionality.
Other implementations in development and in production, such as the MyGateway at the University of Washington, use Windows NT, Microsoft SQL Server as the database application, and ASP as the scripting language. Cornell University has implemented much of their application in Java. ColdFusion, a middle-ware application, could easily be used to perform the scripting language functions. PHP, a free middle-ware application much like ColdFusion, has been suggested as a scripting language solution, and it would work just fine. Of course, the ideal solution for a scripting langauge is one that works across operating systems, database applications, and HTTP servers. Perl and Java presently fit these criteria the best. The other building blocks are pretty much interchangeable.
Present and future My Library systems should consider the use of an XML structure called Resource Description Framework (RDF) [9] to deliver content. During a presentation at OCLC, in Dublin, Ohio, Eric Miller [10] saw MyLibrary@NCState as a system creating content and providing an interface to that content. He saw the content as "channels", the sort of content intended to be delivered by RDF. There are a number of advantages to this perspective. First, since the content is described using a DTD, content is not bound to any particular interface. Since the content is not bound to any particular interface, the content can be made a part of other systems where non-library content is available. Consequently, users could create customized systems including content from news wires, weather services, horoscopes, as well as more scholarly content such as the content provided by libraries. My Netscape implements a service similar to this, but it presently uses Rich Site Summary (RSS) [11] instead of RDF, and only a single RSS file can be incorporated into a user's profile.
Content and staffing issues are the most important issues to be addressed when creating and maintaining a personalized, customizable interface to sets of information resources available from libraries. The importance of these issues can not be understated; they are more important thantechnological issues. It sounds oxymoronic, but computer technology changes continually, and in order to provide consistent service and consistent context the system must be driven by a philosophy of service with people manifesting the concepts articulated by the philosophy. A personalized library interface is manifested by people who have ideas about information service, not by computer functionality. You don't implement a My Library type of service just because you can. You implement a My Library type of service because you believe it will satisfy the information needs of your clientele better than if you had no such service.
To this end a collection management policy must be articulated, written down, communicated to content providers, trusted to the professional judgement of content providers, and finally, reviewed on a regular basis. The collection management policy outlines the goals of the service, what content the service will provide, how the content will be organized and disseminated, how it will be maintained, and who will do the work.
The process of creating this collection management policy is much like the old Indian story where four blind men describe an elephant. One man feels the tail and says the elephant is like a rope. Another man feels the legs and says it's like a tree. The third man feel the elephant's side and says the elephant is like a wall. The last man feels the trunk and believes the elephant is like a snake. In the story, each man has a limited perspective, and in this case the elephant represents the policy and the blind men represent various specialties in librarianship. A cataloger addressing the issues of a collection management policy will have a keen interest in the conceptual organization of the materials. A systems librarian will be strong in articulating the necessary access control mechanisms and technical functionality. A reference librarian will be most interested in the user interface and the system's ability to be personalized for each user and create context. The library administrator will have their eye on monetary costs and staffing issues. Ideally, an individual with more than a working knowledge of each branch of librarianship will be able to create a collection management policy, but that person must also be an excellent leader able to create consensus and motivate groups of people. Such individuals are hard, if not impossible, to find, and in reality, the collection management policy will most likely be articulated by a group of individuals with specific skills and authorized by a single policy-making person.
The types of content made available in a My Library system are a subset of the types of content made available in traditional library services. The defining characteristic of this subset is its digital nature. More likely than not, My Library content will include things like bibliographic databases, electronic journals and texts, dictionaries, encyclopedias, manuals, handbooks, atlases, data sets, Internet resources, links to people, Internet search engines, and bibliographies/"webliographies", etc. All of these things provide the raw material for learning. It will be difficult to provide direct access to a printed book in a My Library system. Yes, you can create pointers to the book, but you will not be able to provide the book itself. Providing access to the book itself is a role of the physical library. On the other hand, creating path finders and finding aids through a My Library system facilitates the identification and location of physical items, but these path finders and finding aids will be digital.
Another type of content will be links to services provided by the traditional library. In other words, there will be links about the library: links to the reference department where the reference department is described, links describing how to renew books, links describing how to obtain an interlibrary load, etc. Finally, a third type of resource available in a My Library setting are links about the community where the library resides. "No man is an island", and information about a library's parent organization, whether that be links to the provost's page or the campus home page, are candidates for My Library content.
Organizing the content provides context, deciding on an organizational scheme will be professionally challenging, and like the entire collection management policy itself, the organizational scheme you use or create will have to be regularly re-examined. Given the increasing amount of specialization in today's world, the use of literary warrant may be the preferred method for creating the organizational scheme. Keep the scheme simple and familiar to your clientele. A hierarchial structure facilitates browsability in a small amount of space, but if the there are too many levels, then simplicity is sacrificed. Similarly, a flat scheme is very easy to browse, but if it gets too long then simplicity is defeated. Existing My Library systems use flat schemes with the number of controlled vocabulary terms ranging from five or six to more than sixty. To give your scheme a head start consider basing iton the hierarchy of your parent organization such as colleges, schools, and departments. Often people will want certain types of information in certain formats. Consequently, supplementing the organizational scheme by format will provide to be more useful. For example, people may very well be looking for demographic data sets in the area of urban planning. A scheme supporting only the "aboutness" of information (urban planning) and not its format (data sets) will not satisfy the needs of the patron. Finally, since information is organic and the needs of your patrons will change, the organizational scheme will have to be organic too. It will need constant revision. Plan to update it regularly.
Digital information is more dynamic than traditional print-based information. This is true because the computer technology used to manifest the information is changing. Newer formats are always making themselves available. Older formats become obsolete. New protocols are being developed, and client software implement the protocols to lesser or greater degree. Consequently, the content of a My Library system, dominated by digital information, requires more maintenance than print-based information systems. Since Web browsers access information through URLs and URLs represent the information in the system, a My Library system should include a URL (link) checker used as the first line of defense making sure links work. There is another type of maintenance though, relevance. My Library systems must be weeded, items must be removed as they become less useful and newer items become more useful. My Library systems are intended to reduce information overload and therefore only the best, not necessarily all, information resources should be available in the system. The collection management policy should therefore define how to remove items as well as add them.
Finally, the collection management policy must articulate who will do the work. Even after a policy is created somebody will have to implement the policy. This takes time. It takes time to identify items for inclusion in the system. It takes time to add them to a database and classify them accordingly. No mater how well your system is designed some people will want to use it and need help doing so. This takes time. Meanwhile present staff already have limitations on their time. Therefore the policy should articulate who is going to do the work and insure these people are given enough time and trust to do the job to the best of their professional ability. If you implement a My Library system you might have to eliminate less used services or explore ways to make current services more efficient.
In summary, a written collection management policy will define how the My Library service will be implemented and provides the framework for its evolution.
My Library systems are nice, but not necessarily enough; they must be truly personalized, and more information needs to be gathered. For example, a number of people like Ian Whyte [12] expressed the need for usage statistics supporting the need of My Library systems. Others such as Art Rhyno [13] believe My Library systems could be used for purposes other than libraries; they could be used for things like My Association, My Lab, or My Community.
This raises the issue of whether or not things like My Library should be subsumed into portal applications of parent organizations. The things such portals might include student grades, tuition payments, and schedules. Undoubtably there would be local news and information about athletic events. Certainly services provided by the campus community would also be available. The difference between these types of information and the types of information provided by libraries is the amount of interpretation necessary to use the information effectively. Grades and tuition payments need little interpretation to understand. The situation is similar with schedules and news. On the other hand, the selection of bibliographic databases or the acquisition of data sets for analysis require more thought by the patron. Put another way, these types of information require a greater amount of context before understanding can be reached. If the information of libraries were relegated to a few lines on portal pages of parent institutions, then the information of libraries would not be as useful as it could be. Using a library and its resources is not like going to a McDonald's fast food restaurant.
A couple of people suggested My Library systems are customizable interfaces and not personalized interfaces. Keith Morgan alluded to Jakob Neilsen and says that My Libraries might benefit from a bit more computer control. Steve Cavrak [14] expressed the desire to have My Library systems turn more into "digital companions." He describes many such features, and some of the more interesting include:
There seems to be a definite interest in services like My Library. With the advent of globally networked computers and a growing service industry, information is seen as an item of increasing value. While people seem to desire to visit libraries less and less, people do seem to desire information more and more. Services like My Library demonstrate one way to satisfy these desires for a larger and larger population. The key to successfully creating a My Library service relies on the abilities of libraries combine their tradition of service with the functions of database technology. Are we up to the challenge?
Eric Lease Morgan
NCSU Libraries
Box 7111
Room 2110
Raleigh, NC
England
27695-7111
Voice: 919/515-4221
Fax: 919/515-3031
Email: eric_morgan@ncsu.edu
URL: <http://www.lib.ncsu.edu/staff/morgan/>
Eric has been a librarian at the NCSU Libraries since the summer of 1991. Previously, he was a medical librarian at the Catawba-Wateree AHEC Library in Lancaster, SC for 3 years. He has a BA in Philosophy from Bethany College, Bethany, WV (1982), and an MIS from Drexel University, Philadelpha, PA (1987)
For citation purposes:
Eric Lease Morgan, "Personalized Library Interfaces",
Exploit Interactive, issue 6, 26th June 2000
URL: <http://www.exploit-lib.org/issue6/libraries/>
[HTML Validation] - [Accessibility check]
|
Issue Home | Editorial | Features | Regular Columns | News and Events | Et cetera | ||
|
| ||
| Go to Top |
A UKOLN Service. Contact Us. Copyright © 1999 |
Last Updated: 26th June 2000 |