Magazine article Computers in Libraries

Expert Systems at ALA; Opinions Vary on the Role of Artificial Intelligence in Libraries

Magazine article Computers in Libraries

Expert Systems at ALA; Opinions Vary on the Role of Artificial Intelligence in Libraries

Article excerpt

This report covers an artificial intelligence session entitled "Librarian Ex Machina 4.0: Planning for Expert Systems," which was held at the ALA annual conference. The session was sponsored by the LITA Artificial Intelligence/Expert Systems Group.

"Planning for the Inevitable" Lloyd Davidson of Northwestern University delivered a paper he prepared with colleagues Peter Schneider, David Levenson, and Kevin Duca entitled Expert Systems in Libraries: Planning for the Inevitable."

The authors are convinced that the use of expert systems is inevitable due to four factors: improvements in computer hardware and software; the increasing expense and scarcity of human experts; the increasing amount and diversity of information; and a change in the mission of libraries, toward more open systems.

The authors believe libraries present an ideal environment for the development of expert systems due to the similarity of the procedures and functions performed m them, and due to the logical nature of those procedures and functions. Expert systems shells will spur development in this environment since they take less time, money, and expertise to create prototypes and finished projects than programming expert systems from the ground up.

Davidson noted some examples of commercial and library expert systems and pointed out several types of cooperative development: sharing of expertise; continuity of projects; and increased interest in expert system projects. Existing cooperative projects could be expanded without the competitive and proprietary interests found in the commercial world.

What can we look forward to in the future? Four areas were mentioned: support for day-to-day operations; enhanced front-ends for databases; improved online catalogs; and improved desktop control over an information universe that has already overloaded us.

In conclusion, Davidson quoted Sam Waters in advising the audience to, "Think big, but start small." Local development coupled with cooperative expert system development promise the best chance for the successful application of this technology in the library field.

Expert Systems:

The Economic, Technological,

and Sociopolitical Aspects UCLA Library and Information Science Professor John Richardson delivered a paper touching on the economic, technological, and sociopolitical aspects of expert systems.

Richardson has reviewed forty-three expert systems in the area of reference. In general, expert systems are still considered a research phenomenon. Most received special funding from external sources such as government agencies and private corporations or from their parent institutions. Financial support ranged from $1,000 to $1.2 million, although most fell into the $3,000 range. The amount of time invested in projects also varied greatly, but the average turned out to be one year's half-time work.

"Knowledge engineering" is where advances will be taking place in the future. Somehow, we must translate our reference expertise into machine-readable form. The problem is not with the technology, but rather in capturing human expertise.

It remains to be seen if expert systems will be as acceptable to the public

Opinions vary on the role of artificial intelligence in libraries as online catalogs. Richardson suggests that the phrase "advisory systems" rather than "expert systems" would be more accurate in describing the function of these systems. There is also a risk of becoming overdependent on the use of expert systems. Jobs won't be lost, but the nature of the jobs will be changed. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.