Denise Duncan
Logistics Management Institute
McLean, Virginia
[email protected]
Though an enduring fiscal crisis imperils the future of all U.S. Navy libraries, one small, but dynamic project demonstrates the resilience of library scientists--even when they are confronted with the specter of library closings.
In the post-cold war era of reduced defense budgets, there are fewer Navy libraries available to support the missions of geographically distributed users. Since 1991, the number of special Navy libraries--medical and academic--has declined by 25%. This number could be headed for an even steeper decline. Call it consolidation, realignment, or right-sizing, the result is that most sailors, marines, and federal workers are no longer within a lunch-time commute of professional library services and a locally-developed collection. Yet, the Navy's voracious appetite for information continues to grow, particularly for technically-oriented reports, journals, and publications.
From September, 1996 through February, 1997, 118 scientists and engineers, aided by 15 librarians at five host libraries, have used the Internet via their desktop browser to search and retrieve material from commercial on-line information databases, including Dialog®. With this system, user transactions are validated against individual permissions, and obligations are posted to individual library accounts. Meanwhile, the librarians negotiate access, licensing, and networking issues on their collective behalf.
While this article will examine the underlying technologies, including the middleware that parses user queries and monitors user activity (statefulness), it is the emerging library business process model, its implications, and potential scalability that the authors find most provocative.
With modest funding from the U.S. Navy Lab/Center Coordinating Group
(NLCCG), the NLCCG Distributed Virtual Library (NVL)
project was initiated in 1995 to improve desktop access and delivery of
scientific and technical information (STI) to the U.S. Navy's Warfare Center
and Naval Research Laboratory community. The project's objective is to
provide its users 24-hour access to high quality STI sources and services
in the Web environment, through library-coordinated acquisition, licensing,
and networking.
User needs were assessed during a requirements analysis phase. The next step was to develop the technology to provide user access and create a delivery mechanism for the desired resources and services. As a parallel activity, negotiations with commercial content providers were undertaken by the LON. A prototype was developed and six months of field tests have recently been completed. The search for funding and sponsorship for full scale development is ongoing.
It should be noted that other U.S. government organizations, notably the Defense Technical Information Center, and the Navy Research Laboratory are working on large-scale digitization projects. Digitization of source collections is not within the scope of the NVL project.
The Librarian of the Navy contracted with the Logistics Management Institute (LMI), to assess user needs. Not surprisingly, field research confirmed that user needs are a moving target, varying as both the availability of information technology and users' perceptions of its usefulness change. However, there is common ground, in this case among scientists and researchers whose demands for accurate, authentic, useful, and well documented technical information could best be met by commercial information providers. This study, conducted between June and December, 1995, isolated the following specific requirements, in order of importance:
The key challenges to providing a variety of customers with remote access
to commercial services lay in controlling the boundaries of the desktop
virtual library environment. For instance, customer transactions have to
be tracked and recorded. Due to low risk tolerance, this could not be done
after the fact, say by monthly vendor billing. Rather, to preserve the
integrity of licensing agreements and preclude financial disaster, each
transaction had to be monitored and verified against permissions in real-time.
Achieving this level of management meant that the system had to maintain control over the customers' interface with external commercial resources. A web-based front-end was needed that, besides guiding user activity, had the capability of trapping customer queries, reconfiguring the syntax as necessary and directing them to the appropriate resource (parsing), as well as returning a result. Given a successful search, product transactions had to verify and record obligations and convey to the customer the variety of document delivery options, such as format and delivery mechanism.
To be useful, the system also had to be capable of processing multiple exchanges per customer per session. In other words, the system had to monitor users' activity as they wander through the virtual library and execute several cost-based transactions, with a variety of providers, each with a different charging scheme. The NVL prototype also needed to offer remote access to a variety of other services, including colleague location, discretionary new publication notification, and local library catalog search.
Given the uncertain future of library funding, system designers were determined to build into the prototype sufficient resolution (granularity) to provide the means for scaling up to a large customer base, and in the process, to achieve lower per-unit-cost.
Three important elements of the NVL prototype had to be engineered by the project team. The first challenge was to create a system based upon individual customer accounts that verified and tracked transactions with information providers and services, such as OCLC FirstSearch. Next, middleware had to be written to parse queries between the NVL web front-end and the provider web-based systems. Finally, the NVL interface had to be created.
In addition to installing the Netscape Commerce Server on a Unix host, software was written to establish the "profiles" that grant or deny customer access to commercial databases, and limit the amount of money spent on any resource. Specificity was required to the level of a single database, display format, and purchase order number(s) for document delivery.
A policy decision was made to enable profiles to be written at both the user and organizational levels, and to allow the individual user's profile to override the permissions in the organization profile. Thus, when the NVL scales to a large user base, profile maintenance will be simplified. An organization (Lab or Warfare Center) will set up default profiles for its organizations. The profile for an individual in one of these organizations can then be modified as necessary, by the librarians on site.
A second piece of software was written to provide a query interface to DIALOG® databases for the scientists/engineers. This software generates the query screen for a given database (Inspec, Compendex Plus, etc.). The underlying tables describe those fields librarians have specified as being most useful, in that specific database, for end-user searching.
NVL customers can execute queries that have been engineered by librarians to be performed by someone other than a professional searcher. This approach to system design results from a philosophy of maximizing utility while minimizing complexity for users. This is an area in which librarians contribute their knowledge of customer needs and capabilities to tailor an interface that the user recognizes and can use easily. A second contribution is made by librarians willing to provide "telereference" services. The NVL includes a utility that allows a librarian to direct her PC's video ouput to an end-user's PC as well. Thus, the expert searcher can perform the more complex searches, with the user "looking over her shoulder", while the user remains in his office. Even with end-user searching, mediation by information professionals remains a vital component in the NVL.
This software was also designed with scaling in mind. Since DIALOG® databases all have the same command syntax, the tables describing them simply list the fields selected for end-user searching, and each field's query mnemonic (e.g., AU for author). Thus, to add one of the 490-plus databases in the DIALOG® system (e.g., GEOREF for earth scientists), one writes a new entry in the configuration file, and then adds the database and limits to the user/organization profiles.
As mentioned, NVL also offers its customers access to OCLC FirstSearch,
using embedded scripts that identify which command a user is coming from,
and decrements that command's account by one search for each search conducted.
Jane's Fighting Ships and Naval Weapons Systems reference databases have
been mounted locally and NVL connects to Current Contents via Goddard Space
Flight Center. Additionally, Faxon and UnCover/Reveal are tapped using
their normal (IP-restriction) controls.
Given a public with Internet access, information providers face a potentially huge, as yet untapped, but widely distributed market, and they are struggling to develop strategies to reach it. In this period of market instability, the opportunity exists for Navy librarians to solve problems troubling both providers and customers by expanding the library's business process model to encompass remote users.
Commercial content and information service providers are experimenting with many Internet customer charging schemes including, pay per use, prepaid accounts, and sample and buy, to name a few. The trade-offs these companies face are not trivial: how much access versus how much security? Responsive service versus managed risk? These are tough choices for business managers in a changing and competitive world.
Further complicating the issues are the challenges of attracting customers, training them to use a sophisticated interface, and gaining their confidence to conduct electronic commerce. Even when successful, it is a costly process to address each customer's unique billing requirements. This also makes forecasting revenue patterns problematic.
The NVL project team found vendors eager to participate, presumably because they, too, wish to extend the library model to the desktop. Having librarians assume a mediator role solves some of their most vexing problems by simplifying customer billing, load-shedding training and risk management, and establishing a stable, predictable revenue stream. In the process, librarians also contribute the expert reference and inter-personal skills that have long characterized the profession.
As mediators, librarians are able to use the leverage of collective
bargaining to negotiate favorable terms and conditions on behalf
of their customers for access and billing. The next challenge is to achieve
harmony in pricing policies to foster an environment where brokers (like
Knight-Ridder Information, Inc.), mediators (like librarians), and end
users (individuals and organizations), can comfortably and efficiently
interact.
D-Lib Magazine's Editor, Amy Friedlander, in the January 1997 issue, challenged the readership to occasionally look beyond the information technologies and our attempts to adapt them to known paradigms, to search for that which is fundamentally different about going digital. With NVL on the cusp of shifting from prototype to operational, the future beckons with interesting opportunities and consequences. We will look at some of them.
At center stage is the changing role of librarians who, in our system, are tackling digitization and archiving issues in one part of the electronic information spectrum and emerging entrepreneurial challenges in another. From one perspective, Navy librarians are pioneers engaged in exciting research to redefine their profession. Glimpsed from another viewpoint, they are being relentlessly pressured by shrinking budgets and an ever increasing demand for new skills.
Some NVL librarians have, by necessity, become adept negotiators. Others have learned to conduct remote training and troubleshoot network problems on the fly. As the project matures, it is reasonable to envision a need for economic skills such as cost modeling, forecasting, and marketing. One wonders if there can be any room left on a librarian's tool belt for even more skills?
One option would be to import these skills by building alliances with professionals from other disciplines, such as computer science and business, in a consortium approach to information acquisition and distribution to users. This tactic presumes that a market exists to support such a consortium. To remain in customer demand, the NVL services must either be unique or their perceived value must exceed what users can negotiate and acquire on their own. This raises an interesting question,
"When the remote customer's library is not the information provider, is its value limited to connecting user and provider?"
If true, there are many alternatives available to the discerning customer.
Further, once the connection is established, the library's presence in
the exchange not only ceases to be beneficial to either party, it could
become an impediment.
This argument may hold water when the resources are free. NVL librarians have chosen to point out free services and resources that complement its other resources, but the system interface merely provides a link, releasing the user with no guarantee of quality or authenticity. However, when the provider's content or service is cost-based, the library's value, as mediator, becomes potentially vast.
Imagine a future where many of the nearly 500,000 personnel in the Naval Service scramble, without benefit of librarian mediation, to negotiate individually and in small groups for desktop access to cost-based information. Consider the percent of an organization's total funds, time, and effort that will be expended in replicating those services provided by the NVL prototype. Imagine the impact on the quality of work and efficiency of production. Try to predict how much will it cost the Navy, in lost savings, not to negotiate fees as a united customer base. Each of these factors is mirrored on the provider's side of the equation.
Our economic analysis seems to confirm that the value of mediation in the realm of electronic information is so enormous that the market will force its creation. Librarians are uniquely qualified to fill this void. Yet, our experience tells us that the required skill set is too diverse to come from only one discipline. We conclude that the future business model for Navy libraries will best be executed by a consortium of people, including librarians, business analysts, and managers, each contributing unique skills to an organization that is keen to increase its scope and influence.
Among the consequences of going digital is the opportunity for librarians to build alliances with other disciplines, especially those in the business world and computer science, as we shape the virtual component of our libraries to mediate on behalf of customers who never set foot in the door, and providers who service a world-wide market. The potential value of mediation, to remote customers and information providers, points to a future where the base of support for some libraries could become enormous. Leveraging that support into real budgets will require an entrepreneurial mindset and a willingness to embrace change. In the case of the NVL project, the impetus for that change emerged from adversity.
The NVL design is elegant technically (no wasted effort). It more than adequately satisfies its charter and is scaleable without losing integrity. Upon reflection, it works because it was well engineered, and just as importantly, it was fortuitously timed to be in synch with its environment. The prototype did not encounter apathetic or untrainable users, a lack of broker support, or librarian resistance, any of which would have been deadly. NVL is the product of a crisis in the Navy's library community that has, fittingly, challenged us to sink or swim.
hdl:cnri.dlib/march97-norris
