INFOMINE--A model Web-based academic virtual library

Abstract

INFOMINE is a World Wide Web virtual library which provides indexing, annotations, and links to Internet resources of scholarly use to the academic community. Thousands of Internet resources covering most major disciplines are present. A custom hypertext markup language (HTML) converter and database manager allow librarians with subject expertise but no HTML experience (or those with HTML knowledge but no time to maintain several HTML indexes) to contribute to and maintain INFOMINE in a time-efficient manner. INFOMINE was developed by the Library of the University of California, Riverside. It was one of the first Web-based, academic virtual libraries, as well as one of the first to develop a system combining the advantages of the hypertext and multimedia capabilities of the Web with the organizational and retrieval functions of a database manager. Some of INFOMINE's important features and services, which remain unique among Internet resource collections, are described.

* Introduction and History

INFOMINE is a unique Web resource featuring well organized access to a substantial number of important research and educational tools on. the Internet. INFOMINE is notable for its collection of close to five thousand annotated and indexed records with links to selected, university-level resources in most major academic disciplines. Information in INFOMINE is easy to find given the multiplicity of access points provided. It has received, on average, over twenty thousand accesses per week during the last fifteen months. In addition, INFOMINE provides simple and streamlined capabilities for adding resources (e.g., knowledge of HTML is not required) as well as essential maintenance functions (e.g., a URL Checker ensures that links are good). This article discusses INFOMINE as both a large Internet resource collection, commonly known as a "virtual" or "digital" library, and the custom hypertext database management system, with HTML converter, that has made the collection easy to build, maintain and use. INFOMINE (figure 1) can be accessed at http://lib-www.ucr.edu. (Figure 1 omitted)

INFOMINE began over two years ago and was developed by the Library of the University of California, Riverside. It was one of the first Web-based, academic virtual libraries, as well as one of the first to develop a system combining the advantages of the hypertext and multimedia capabilities of the Web with the organizational and retrieval functions of a database manager.

Most of its important features and services, described in-depth below, remain unique among Internet resource collections.

* INFOMINE in Depth

The INFOMINE Collection

INFOMINE contains close to five thousand records. Among these are substantive databases, guides to the Internet and other electronic resources for most disciplines, textbooks, conference proceedings, and journals. The life sciences INFOMINE alone provides interactive access to nearly three hundred databases.

Separate virtual collections or INFOMINEs exist in most major areas of university-level research and educational interests. These include collections in: (a) Biological, Agricultural and Medical Resources; (b) Government Information Resources; (c) Social Sciences and Humanities (included here as well are General Reference, Business, Education, and Library/Information Studies related resources); (d) Physical Sciences, Engineering, Computer Science and Mathematics; (e) Internet Enabling Tools (e.g., help, tutorials, navigators of assistance in Internet usage); (f) Maps and Geographic Information Systems; (g) Visual and Performing Arts; and, (h) Instructional Resources (in all disciplines) on the Internet. These are listed in order of comprehensiveness. The first two collections contain fifteen hundred and thirteen hundred resources respectively. The others are strong and compare favorably with collections at other large university systems surveyed. All files are growing rapidly, reflecting the growth of the Internet, and represent well-organized collections of selected, high-quality resources.

The INFOMINE System

INFOMINE for Users

Among the contributions of INFOMINE is the essential enrichment or "value added" service, as mentioned, of providing annotations as well as in-depth indexing terminology for each record. This greatly helps faculty and students to quickly retrieve a focused results set, examine the relevance of individual records and then choose among them immediately prior to accessing thus saving considerable time.

In addition to providing search capabilities on conventional subject, keyword and title terms, INFOMINE allows retrieval on terms such as "comprehensive," "reference resources," "subject guides," "virtual libraries," and "searchable databases," among other terms, which indicate the depth and scope and/or pathfinding nature of specific resources. These qualities and others are often indicated, as well; in the annotation. Also emphasized is the inclusion of important subject guides to Internet, as well as to other electronic and print format resources in most major disciplines. Concerns regarding resource comprehensiveness, quality, and general usefulness from an academic perspective guide all INFOMINE resource selection activities.

INFOMINE provides a great number of access points, through both browse (What's New, Title, Table of Contents, Subject, Keyword, Title, hyperlinked indexing) (figure 2) and search (title, subject and keyword) (figure 3) modes. (Figures 2 and 3 omitted) Search mode allows the user to quickly retrieve among the collection on the chosen subject(s). Nested, Boolean searching capabilities are featured. Search results come back in the form of dynamically created Web pages which are custom HTML documents created on the fly and containing the results set unique to each search (figure 4). (Figure 4 omitted) By way of contrast, many conventional Web virtual libraries provide minimal searching and/or depend on displays of static lists of resources. In INFOMINE, each results set record, in addition, features indexing terms that are viewable and in hyperlink form and, when clicked upon, allow further broadening or narrowing of the search as desired (figure 5). (Figure 5 omitted)

Noteworthy for those browsing is the Table of Contents (figure 6). (Figure 6 omitted) In this, all subjects are listed with the titles of all resources on a subject filing under the given subject term. For those who know which resource they want to use, clicking on the title will open it up for use. For those needing a little more information, clicking on a subject will bring up the title plus annotation for all resources listed under that subject.

In-depth description and indexing, careful selection, a considerable number of options in browsing/searching (access points), and ample help within INFOMINE combine to assist the academic community in discovering and using important Internet tools.

INFOMINE for Indexers--Content Development and HTML Document Creation Made Easy

Several features and tools for resource description and indexing make INFOMINE attractive among those contributing to the building of the files. They combine to make INFOMINE records easy to add, edit, and maintain. Adding URLs (uniform resource locators or resource addresses), titles, subjects, keywords and annotations is simple and straightforward. Often, within the Windows or Linux environments (among others), adding a record is simply a matter of cutting and pasting.

It is crucial to note that in adding or editing, one does not need to know HTML. All new records are converted into this format automatically, when they are displayed, by INFOMINE. As a result, participants with subject knowledge, but no HTML experience, need not be challenged to learn yet another system perceived as complicated.

Moreover, from our central adder document (figure 7), data entered for each record is automatically converted to HTML in each of six dynamically created INFOMINE HTML documents at the time these documents are requested for viewing (i.e., the search results set as well as the following indexes: Titles, Subjects, Keywords, Table of Contents, Date-What's New, and, soon, Author). (Figure 7 omitted) By way of contrast, in virtual libraries or subject guides made up of static HTML resource lists/documents, data representing each individual record would have to be manually entered in each of the separate HTML indexing documents. INFOMINE's ability to allow the adding of record data into one document and have this data then be automatically expressed in several HTML documents (again, results set pages and indexes) saves substantial amounts of time that would otherwise go into manual HTML work. For this reason, INFOMINE has provided a time-efficient means of building a substantial virtual library for both those with and without significant HTML experience.

Another helpful feature that is indispensable in maintaining large collections of links is a program that will automatically check and flag, at specified intervals, resources which have changed locations. INFOMINE's URL Checker does just this. It is important to note in this regard, for example, that 5 percent of the resources in the life sciences component of INFOMINE have moved over the last ten-month period checked (the majority of these, though, have simply moved to a slightly different place on the same server, thus making them easy to relocate). In addition, a slight variant of the URL Checker has been employed to ensure that the resource to be added is not a duplicate (important with the current size of most of the files).

As INFOMINE has grown, various paths for future development have presented themselves. For instance, we have started exploring the development of networks of cooperating selectors/indexers not only at UCR but at other UC campuses as well, INFOMINE presents significant opportunities for efficient, multi-campus, shared Internet resource collection. Such efforts would result in a significant reduction in redundant collecting efforts at each individual campus. An informal pilot project finished during the summer of 1995 has been successful at outlining many of the issues and concerns that would be a part of such a project. We are also examining the need for more uniform indexing languages for the various INFOMINEs. To this end, we have begun using Library of Congress Subject Headings. In addition, shared indexing terminology among INFOMINEs already exists, providing conventions and protocols for describing document types and approaches to geographic descriptors. Another ongoing project includes the modification of the INFOMINE management system so that both the data files and management programs could be distributed to participating campuses/organizations.

INFOMINE System Technical Details Summary

* The INFOMINE Web server is a 90 MHz Pentium running Apache (httpd server, v. 1.0) on Linux (kernel v. 1.3.45), a form of the UNIX operating system for PCs. Apache and Linux were chosen because of their price (free) and performance. Linux is now believed to be the best operating system, given its overall performance, for the i386 architecture family of PCs (386, 486 and pentium machines).

* INFOMINE, though optimized for the current version of Netscape, can be accessed by all popular Web browsers.

* INFOMINE uses mSQL (mini SQL), v.1.0.10, as a database management system/server to facilitate indexing and the storage/retrieval of records. This standard SQL-based system will allow INFOMINE to grow substantially in the future. mSQL is known to compile on many popular UNIX systems, such as Solaris and SunOS, OSF/1, among others.

* INFOMINE could be served from very powerful, high-performance workstations when the need for greater performance presents itself.

* INFOMINE utilizes the C++ programming language for all user-interface functions (e.g., searching, displaying the table of contents, adding/editing resources, etc.). The PERL programming language and C++ are used for INFOMINE system maintenance functions/programs.

* For INFOMINE contributors, a Web forms-based record adder/editor is provided which allows easy adding/editing/deleting of multiple HTML documents as appropriate. In adding and editing, INFOMINE features a built-in automatic HTML conversion feature which eliminates the need for contributors to acquire HTML skills or make additions/changes from more than one HTML document (attractive to those with HTML experience).

* Easy, efficient maintenance of a large digital library depends on having an automatic way of checking on the working "linkability" of the records included. INFOMINE provides this in the form of its URL Checker. This checks for and then flags records with links that will not open for later annual verification by databases editors.

* Conclusion and Project Personnel

It is important to note in ending that many of us who directly provide Internet access to faculty and students have found that expectations for robotic and other Internet navigational or finding tools which provide very minimal human input have not been met. In comparison, the INFOMINE virtual library is an efficient and academically focused organizing tool which joins the vast, traditional experience of our profession in organizing information with the Internet in order to help create intelligent order among and access to high-quality, well-selected, and annotated electronic resources.

The INFOMINE project development team consists of Steve Mitchell (UCR Sciences Librarian and INFOMINE Co-Coordinator; smitch(at)ucrac1.ucr.edu); Margaret Mooney (UCR Librarian, Head of the Government Publications Department, INFOMINE Co-Coordinator; mmooney(at)ucrac1.ucr.edu); Carlos Rodriguez (UCR Sciences Librarian, INFOMINE Graphics/HTML Coordinator and Physical Sciences Resources Facilitator); George Daswani (Programmer/Systems Analyst, UCR Library).

Steve Mitchell is the Sciences Librarian and Margaret Mooney is the Head of the Government Publications Department at the University of California-Riverside.