Printer Friendly

ASIS mid-year: May 13-17, 1990 Fort Lauderdale, Florida.

ASIS Mid-Year May 13-17, 1990 Fort Lauderdale, Florida

Sandwiched between two entertaining keynoters, a few speakers read their contributed papers, a few more were invited by Special Interest Groups to read theirs, and a lot of registrants sunned and swam poolside and beachside. In the first four sessions I attended, three contributed paper authors never showed up for presentations, one paper was read by a friend of the author, and a substitute chair let his speakers introduce themselves. That's why I now have the best tan I've had in years, and it's only May.

The meeting had other problems. The theme for this Mid-year was "Micro-computing in the 1990's: Unlocking the Power." The subject of microcomputing for a meeting is already old-hat. And how many of us really know enough to project and predict about any subject. The subject thus becomes a great excuse for submitting a paper on projects not only in progress but even in contemplation.

It is dangerous to have meetings at resorts. Are abstracts of proposed papers perhaps submitted because the location is so attractive, then accepted by the program committee before the author has permission to attend the meeting? Is permission then not granted? Whatever the reasons, ASIS had to deal with a lot of irresponsibility. Is there some intangible that has fostered this irresponsibily? ASIS ought to find some tangible approach to avoid the problem. It just doesn't look good.

Keynoter #1-John Sifonis

Now to the entertaining keynoters. The first was stimulating because of and despite of what he didn't say and should have. John Sifonis of Temple, Barker & Sloane spoke directly to the theme of the meeting. He had been part of a "1990's Program Study," a $5 million project at MIT begun in 1984, sponsored by eight corporations. The study was to predict the 1990's impact of technology on corporations. It turned out the study concluded that organizational behavior would be a most important factor and that putting technology in place would have a great effect on the organizational culture of a corporation.

Sifonis spoke of vision, faith, and a dream a la Martin Luther King and how more important these characteristics were than Critical Impact Factors. The audience, including me, loved that. We loved his next point too. There must be a commitment of time when one wants to affect real change and realized the full affect of technology. But--the next points he made didn't quite follow. He reviewed the eight corporations sponsoring the impact study and the dream and vision spirits seemed to disappear in thin air. General Motors was a sponsor because it was having trouble surviving. Bell South was very worried about MCI. Kodak simply had to change with the times and think of themselves as image processors or they would have had it, etc.

It would be nice to believe these and other corporations have vision these days but I'm skeptical. These weren't dreams. They were pressure points. And how about that long term commitment? Those corporations will want immediate results. I'll bet they bridle when they find out they'll only succeed in the long run. It's pretty obvious today that our problems are acute because we don't invest in the long run as the Japanese have been doing. I well remember the post World War II days when dreams and up front large chunks of money were available. Dreams and large up-front investments are not what accountants and MBA-graduate CEO's or CIO's are made up of.

This talk frankly confused me. Would microcomputers and their impact on changing organizational behavior rid us of non-dreamer management?

Sifonis ended his talk with a list of trends--there will be less mainframes and more PC's; expect lots of investment in the education and medical areas; software will sell well; white collar computing will increase; and so will online database revenue. Oh, well..... and ho hum. He also managed in the middle of his talk to add to my list of "funny" information words such as "disambiguating," "deduplicating" (now "deduping"), and "disintermediating." Sifonis "automates" after which he "informates". That's really not that bad; it does make some sense. During the sessions I also heard one speaker refer to "persuasive" locations. Try to figure that one out.

Keynoter #2-Alan Paller

Allan Paller was excellent. He was a major reason I came to the meeting, and he didn't disappoint. He was unfortunately left hanging at the end of his talk with his last slide stating. "ANY QUESTIONS?" as the session was abruptly closed because of lack of time. This was all due to excessive introductions.

Mr. Paller is an Executive Information System (EIS) expert. He's just written a book on EIS and is editor of EIS Conference Report a monthly newsletter on the subject. I recently did a rather extensive paper on EIS and found Alan Paller quoted a great deal in the journals dealing with EIS, journals such as Computer World, etc.

EIS has become a common acronym in only the last year and a half or so. It has been hard to find a definition, but it seems the characteristics that differentiate it from Decision Support system (DSS) are that the emphasis is on hitting the top executives with vital information and intergrating external competitive information with the usual MIS internal sources. Early on, Alan Paller stated, and well that he did, that he had defined EIS in his own way. He emphasized EIS should satisfy the operational executives as well as the very top dogs, and therein he seemed to be regressing to DSS because those are the people which support systems are supposed to be already reaching.

More importantly, he emphasized the need for human experts--yes you heard it right--human experts to scan, analyze, and integrate the external information. I kept hearing at this meeting about expert system software, i.e., expert system shells which need filling. How exciting to hear that shell technology required human fillers--and that the humans are as important as the technology. I hate to start up anew the old argument about whether it is necessary that the important scientist or librarian be educated in the domain of his information world, but can you imagine the expertise and responsibility of the human integrator Alan Paller is describing--he'd better have plenty of domain smarts.

Alan Paller, in the most important part of his talk, sounded as though he was instituting a Management By Objectives system. Remember MBO? The information systems he wants instituted will not help management make decisions but will ensure that policies are implemented. The systems are not set up to deal with problems that have already occurred but to ones that need heading off. The information fed to the executives must meet a specific objective. This was perhaps Paller's most important message.

He went on the warn against just presenting pretty pictures and graphs and suggested more text. He also advised underpromising and overdeveloping. His procedure outline for developing an EIS was: start with a real problem; skip the design phase; deliver a prototype (delivery being the keyword); and redo the prototype, thereby beginning a never-ending development of a practical system--he called it a journey. In truth, I don't know of any successful information system in any field or domain that hasn't progressed that way. But, be sure you allow a lot of money for including the human experts, expensive as they may be.

The Rest of the Meeting

Most of the papers at this meeting were of secondary interet to me so my review of papers other than the keynotes is restricted to some brief notes and mostly conclusions. Kathy Miller of Oak Ridge National Laboratory in Tennessee spoke about expert system activities, specifically about selecting a small expert system and introducing it into a research environment. This paper was almost as much fun as a Miss America Pageant. The almost was a big almost, software not being quite as attractive as bathing beauties. In her project to find the most practical software shells for researchers to use, the 1988 project evaluated 20 different software packages to produce five semi-finalists: 1st Class; Insight 2+ (now called Level 5); EXSYS; VP-Expert; and Intelligence Compiler. The last two are co-winners but as the paper progressed it seemed that VP-Expert was the real winner. It is expected that engineers who pick up the expert system shell and fill it with something useful will produce a 15 to 1 return investments, and there are practical applications that seem to prove it out.

Helen Pfuderer also of Oak Ridge National Laboratory described the development of a practical expert system, "Smartfrom" (including hypertext) for an environmental system.

We're talking real practice here--software chosen specifically for easy use and inexpensive as well. The project was based on the 1985 Dupont experience of 1985.

Helen Pfuderer also of Oak Ridge National Laboratory described the development of a practical expert system, "Smartfrom" (including hypertext) for an environmental system.

William Cooper of the School of Library and Information Studies at the University of California, Berkeley, delivered an interesting little provocative treatise which Hal Borko, recently retired from UCLA found a little difficult to accept. Cooper's paper was titled "Maximum Entropy Computations for Probablistic Information Retrieval On a Microcomputer." For those of you interested in probabilistic information retrieval read on. Otherwise this report is finished.

Cooper issued an outline of his logic to participants. I quote his last three points. I think they deserve cogitation.

"It has been demonstrated that maximum entropy computations for requests containing up to five terms or term clusters can be completed in a few seconds on a fast microcomputer.

"This would be sufficient to allow the use of maximum entropy methods in some microcomputer IR programs or user interfaces."

"One possibility is 'Probaboolistic IR', in which the retrieval system accepts taceted Boolean requests containing no more than five facets. The Boolean request is optionally transformable into a probabilistically weighted request."
COPYRIGHT 1990 Information Today, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1990 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Brenner, Ev
Publication:Information Today
Date:Jun 1, 1990
Words:1652
Previous Article:CALR training: costs, sources, & content.
Next Article:Talking Newspapers voice services help newspapers provide information on demand.
Topics:

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters