Student Advisory
Research from Home
Dean's Desk
Librarians Who Read
Collection CoNextion
Why Is It?
Library Hours

Editors' Page
Previous EDGE
Subscribe
Home (This Issue)

Dean's Desk: Books that talk to each other
Ralph Alberico, Dean of Libraries and Educational  Technology 

Marvin Minsky, MIT professor and artificial intelligence researcher, predicted that in the future people would be surprised that there were once libraries where the books didn't talk to each other.  In some sense the books and other carriers of ideas in libraries have always talked to one another.  Or, more accurately, the thinkers and doers who have recorded their findings and ideas have always talked to one another.  The system of scholarly communication that has evolved over hundreds of years is an ongoing conversation that transcends time and distance. The references and cross references, bibliographies, indexes, notes and hyperlinks associated with scholarly publication are all examples of the ways in which the books and other media talk to one another. 

Online catalogs and indexes
On another level, the indexes and abstracts, catalogs and databases that libraries support are examples of the ways in which libraries talk to their communities.  The Research Tools section of our web site, including LEO is our way of telling you about what we offer in the library.  LEO works best when it is telling you about the books on our shelves, less effectively when it comes to the articles in our journals and even less for information on the internet.  So, in addition to LEO, the library offers hundreds of indexing and abstracting services, mostly on the web nowadays in the form of databases.  In fact, a good deal of the work that goes on in our library is related to development, acquisition, user support and maintenance for LEO and the other research databases that tell students and faculty about information that has passed our standards of quality.  This effort permeates the library; almost everyone who works in the library is involved to one degree or another in managing our research databases. 

The library community was quick to automate the tools for providing author, title and subject access to books and other sources of information.  Card catalogs and print editions of subject indexes morphed into online databases and provided many, many more points of access to library holdings than paper possibly could.  Library systems and research databases were easy to invent based on existing approaches, widely accepted protocols and the limitations of the technology available when they were invented.  At the core of most of the research databases was a printed index.  Things happened fast.  In a few years most library card catalogs went online so did the major indexing and abstracting services.  Publishers of print reference tools dropped their print offerings in favor of online access. 

These gains have not come without a price.  Even as the amount of good quality information increases and possibilities for connection grow richer, searching for good information has become more confusing, diffused and frustrating than it once was.  There are so many options now that simply choosing a place to begin has become a difficult task. The library offers several hundred online databases and electronic journals, each with its own structure, subject vocabulary and user interface.  Even for experienced researchers, the system isn't easy to use.  

Cross reference systems
The library is making steps toward a more seamless information system.  Just over the horizon is the possibility of a network of interconnected articles, books and author, title, and subject access tools.  The next phase of development will be more difficult.  The task of providing the academic community with a consistent and coherent view of the high quality information the library offers is a major challenge.   Another challenge involves providing an intuitive way of navigating across large collections of articles and books.   Right now researchers have to spend too much attention on looking for information.   

I am convinced that emerging technologies and new standards that weren't there when the first generation library systems were invented will make it possible to meet these challenges.  In the library, we are just beginning to experiment with cross-referencing and linking items in our collections.  Eventually, there will be widely accepted norms for representing and connecting pieces of information and common search interfaces for the databases offered on the library's web site.   

Online catalogs and indexes were easier to develop than systems that capture the natural connections that arise from within articles and books, web sites, research communities and work groups.  Until recently, there weren't enough articles and books available to make online access to cited references feasible. Nor did the technology support the kind of indexing and cross-referencing capability that would be needed.  When the first generation library systems were invented, content analysis of books and articles was not easily possible.  The idea of an information system that develops a relationship with its users was almost science fiction.  The brute force computer power to do the kind of profiling and pattern matching that is now so routine was not there.  Things have changed. With the advent of the web, we have seen an explosion of systems that allow researchers to follow natural connections in a hypertext environment.  

Most of the library's online journals (e.g. Science, Annual Reviews, IDEAL (Academic Press) allow readers to follow links from article references to other journals that we offer in electronic format.  We now have a database, Web of Science®, that allows researchers to explore connections between science and social science research articles and to connect to any article that we have licensed or purchased.  Soon most of our index and abstract databases will connect directly to articles that the library has licensed.  Progress is being made in improving subject access systems at the same time that new ways of representing and organizing information are emerging.   Students and faculty can now also explore connections among the items in our collections -- jumping from an index to an article, from an article to another article, from a citation to a group of articles, from a collection of citations to a search profile -- all of this with existing tools. 

Eventually we will be able to combine the highly structured access offered by our catalog and research databases with the natural connections that arise within collections of documents.  On one front we are working toward an integration of the research tools that will make it easier for students and faculty to search for information and to find items in our collections.  On another front we are acquiring and developing tools that will allow us to connect articles, books, and information collections to one another.   The books and articles don't talk to one another yet, at least in the immediate and automatic sense that Minsky had in mind but they are beginning to introduce themselves.

E-mail comments and questions to:
clarkeke@jmu.edu

Copyright © 2001. JMU Libraries.
All rights reserved
.