The opportunities granted by the Internet do not, however, come without corresponding limitations. The electronic mail, file transfer, and remote logon protocols that made the Internet successful assumed that the user knew in advance what information was available and how to gain access to that information. Although various directories and listings of sources of information and their electronic addresses have circulated formally and informally since the Internet's infancy, they have acted only as a partial solution to providing effective access to network resources.
The more complete solution lies in the development of sophisticated navigational tools to guide the user through the maze of the Internet. The key word in the preceding sentence is sophisticated. Although several navigational tools have gained popularity, none of them is more than a simple, preliminary, experimental approach to a complex, longterm problem. Mosaic, Yahoo, and other web browsers provide visually satisfying interfaces to Internet functions via WWW (World Wide Web), but add nothing tangible in terms of effective utilization of network resources.
Recently there has been agreement and much discussion among Internet users of the lack of organization and the poor quality of the indexing of the vast amount of information available on the Internet [2]. Snyder [3] states there will be "too much information too poorly indexed [on the Internet] to be of any use to anybody."
Response to the Internet and its potential has generally been more emotional than analytical. A typical remark comes from David Churbuck [4], writing in Forbes: "Fulltext retrieval, still very limited at this point, is around the corner. When it comes, the local library as we know it all but disappears. In lieu of librarians we will have programmers and database experts." The naivete of this and similar statements is initially disturbing, but implies an unexpected compliment to the library profession. Librarians have done their jobs so effectively, have made their libraries so usable that educated, reasonably intelligent individuals, such as Churbuck [4], fail to understand that any work has been done at all. This is the epitome of the transparent, userfriendly interface. The reality, of course, is that professional librarians are database experts at a level that is incomprehensible to the mainstream of computer science. Library and information science has historically concentrated on intellectual access and usability. The former is typically treated as irrelevant to the design of computer databases. The latter is addressed primarily in a mechanistic manner that concentrates on hardware and software features at the expense of attention to human characteristics.
The current methods of retrieval on the Internet are more limited than the highly structured commercially produced proprietary data bases often available at libraries. The difference between keyword searching and controlled vocabulary searching is the distinction between mechanistic access to isolated terms and intellectual access to unified concepts. This is the key to the utilization of the Internet.
Since libraries, specifically librarians, have historically been involved in the organization and retrieval of information, the terms digital library and library without walls have emerged to exemplify the library's active participation in the selection, creation, and representation of electronic resources. Although there is no agreed upon definition of the digital library, many librarians refer to the new library as a "gateway" to electronic resources, not as a collection of information resources [5]. Richard Rockwell [6] defined the gateway library as "an integrated and organized means of electronic access to dispersed information resources." In this sense "the digital library is a step toward a gateway library, but lacks the organization and retrieval mechanisms that would ultimately be available" [5].
Library and information professionals are expanding their role in the digital library toward the move to the gateway library with the inclusion of the scholarly and scientific communication process. This expanded role has been termed, "knowledge management" and allows scholars, scientists, and library and information professionals to utilize the tools developed by computing and telecommunications experts to "embrace[s] the entire information transfer cycle, from the creation, structuring, and representation of information to its dissemination and use" [7].
Engineering researchers often need the most current information in their discipline and the Internet is often the best source of the most current engineering resources. Thomas P. Dowling [11] identifies a plethora of engineering resources available on the Internet, with no systematic method of retrieval available for these resources. The purpose of this project is to identify electronic resources that could be of value to engineers and to represent these resources in a manner that enables engineers to make timely, informed decisions about the usefulness of the resources. The specific objectives of the first phase of this project include: 1) the development of a selection policy for engineering electronic resources; 2) the identification of electronic resources of interest to engineers, as defined by the selection policy; and 3) the creation of abstracts for these electronic resources that will include at least two hyperlinks to other related electronic resources.
The DU project team was contracted by Ei, Inc. to evaluate and determine if the 2,000 electronic sites already included in the International House warranted inclusion, identify 1,000 new sites to include in the International House, and abstract each of the sites to be included in the International House. Therefore, it was necessary to develop selection criteria for inclusion in the Ei Village International House prior to identifying or abstracting the electronic engineering resources. If sites did not match the selection criteria or were no longer viable, they were deleted from the existing list of sites or not selected for inclusion in the Ei Village International House as a new site.
When visiting a site to evaluate for inclusion in the International House, the site was deleted from the list if the address was no longer viable. It is common to access a site on the Internet one day and the next be given a new address or more frustrating yet, receive a "file not found" message. When URLs change there are often references to the new addresses. This takes much time to continually update. This is why the documents were downloaded and the date of downloading was documented. Once a site was accessed, it was determined if the information was reliable. Identifying reliable sites was often difficult. The DU project team used the following criteria to evaluate a new site for inclusion in the Ei Village International House:
-- Whose site is it? Is it a person, a company, government?
-- Who is the author, if this is an article? Is the author qualified to write about this subject?
-- How does the information compare with that of other sources in the field?
-- Is the web site updated and revised frequently?
It is often difficult to tell the origin of a home page, i.e., the author, last update, etc. This information should be available to warrant inclusion in the Ei Village International House.
In addition to the site's reliability, the project team's general selection criteria for including existing and new sites in the Ei Village International House are: sites should benefit engineers in every aspect of professional work, including engineering proper; news about all countries; travel information; government information; and business relations within and between countries.
Sites should contain substantive information, not merely links to other sites. It is difficult to retrieve web sites that contain substantive information. Searching for web sites often retrieved irrelevant sources, e.g., advertisements for departments, individuals, corporations, associations, societies. Web sites each had many links to other corporations, associations, societies, etc. They were primarily links to more links -- metasites. The web searches retrieved advertising sites. It was often difficult to establish a site as a single web site since it can be composed of multiple files and hypertext links to other documents and document sites. In general, metasites, sites that provide pointers to other sites, were not considered for selection unless they provided unique information not found elsewhere.
Sites should be clear as to content and navigability. Sites that are international in scope and contain discrete pages for individual countries may be treated both as a general global site and as individual countries, depending on the quality of the information.
Abstracting Internet sites is a challenge because every site is organized differently. It is important to access a site and its content quickly. Ei, Inc. required indicative abstracts. The abstracts contained general statements about the sites, with no attempt to report the actual content of the site. Ei, Inc. also stipulated the inclusion of two hyperlinks or hotlinks to relevant sections of each abstracted site.
The DU project team believes it is possible to utilize the tools developed by computing and telecommunications experts to accommodate the structuring and representation of information to its dissemination and use (Lucier's 1990 definition of knowledge representation) [7] in order to provide a gateway library. First, the existence of the information on the Internet must be identified. After the information is identified, it must be evaluated in terms of the users and existing resources, indexed, abstracted, and represented in a form that can be logically searched on a web site and incorporated into other electronic retrieval systems. If this type of information is linked to other electronic retrieval systems, the information will have to be continually monitored and updated.
The project includes no specific examination of user behavior in relationship to an Internet navigational system based on selection criteria, abstracting and indexing language principles, and the structure and representation of subjects and categories included in the web site. A logical next step would be to examine the distinctions among user approaches to such a system and user approaches to existing navigational tools.
Testing may be conducted to compare use of the knowledge management approach to alternative, available approaches to identify and retrieve electronic engineering resources. Comparative analysis will be achieved by constructing a set of typical informationseeking profiles and executing each profile using each of several navigational tools, including the knowledge management approach accessible through the Ei, Inc. International Village. Success will be measured in terms of the total time required to achieve an acceptable result, the number of steps required to achieve an acceptable result, the volume of information produced by each search and the relevance of the information retrieved.
Note: This research project was partially funded by the 1995 American Library Association Carroll Preston Baber Research Grant and a 1995 University of Missouri Alumni Association and will be published in Finding Common Ground: Creating the Library of the Future Without Diminishing the Library of the Past, The New Library Series, no. 3, Neal-Schuman Pub., Forthcoming Winter 1997-98.