Corrections or additions?

This article written by Melinda Sherwood was published in U.S. 1

Newspaper on April 21, 1999. All rights reserved.

Meet HAL, He Wants You to Share Your Data

Remember HAL, the omnipresent computer in Stanley Kubrick’s

film 2001 that was so perfectly human that it became perfectly malevolent?

Although still in their infancy, similar "thinking" computers

are now becoming an important tool in business, as a result of software

that enables computers to "learn."

"Knowledge-enabling" technology, says Michael Charney,

vice president of Corporate Knowledge Services at Molloy Group Inc.

in Parsippany, is going to determine which businesses survive in an

era of internet-driven commerce. "Consumers are much more empowered

today," he says, pointing to how the internet is both guiding

sales and providing customer support services. "If you’re looking

to make customers happier," he says, "an intelligent technology

is a way to do that."

Charney will lead a discussion on "Preparing for Knowledge Management

in the E-commerce World" at the PSE&G Training Center at 234 Pierson

Avenue in Edison on Tuesday, April 27, at 5 p.m. Cost: $30. The Knowledge

Management Forum is sponsored by Technology New Jersey. Call 609-414-4444.

Charney holds a BA in psychology from Berkeley, Class of 1980, and

has spent the past 20 years working with technology, first as a systems

analyst and later as an instructor of new software. He received an

MA in English from William Paterson and published his own fiction

and non-fiction works in both college journals and high school text

books. "I am fascinated with language and the way the mind works,"

he says, "so when I had an opportunity to help design new models

for implementing knowledge, and learn about the technology, I felt

very fortunate."

At Molloy Group, Charney draws on his own knowledge

and experience in psychology, linguistics, teaching, and computer

science to create a methodology for implementing knowledge system

products — specifically, a proprietary piece of software called

a "cognitive processor." This, he explains, is what enables

computers to model human thinking when tapping into traditional databases.

"Until now," he says, "computer `thinking’ has been rather

primitive." Decision trees, he explains, enable computers to come

up with simple answers based on "yes" or "no" questioning,

a model that is both slow and limited. Case-based reasoning allows

computers to superimpose queries on paradigms until an appropriate

match and, consequently, solution to a problem is found. "This

is good for static knowledge," Charney says, "but if you have

knowledge that changes frequently, you have to keep changing and updating

the models."

The new technology, called a "neural network," enables computers

to make sophisticated associations when dealing with a query. "If

you punch in the word `sun,’" Charney says, "you’ll also get

moon, and everything related to it." The process is organic; it

is about making connections, similar to the way in which the creative

human mind works.

Charney sees the new technology becoming an essential tool in businesses

that rely on customer service and support. A customer service representative

dealing with a billing problem, for example, would be able to tap

into the collective experience of anyone who has ever dealt with a

similar problem and could retrieve a sophisticated answer quickly.

The advantages are obvious, says Charney. "The website or telephone

support center that allows me to do my business quickly and efficiently,"

he says, "is the one that’s going to get my business."

Businesses that benefit from the new technology should prepare to

undergo cultural changes, not just technological changes. Charney

offers the following advice:

Encourage knowledge sharing. Businesses should provide

incentives to employees for careful documentation and sharing of information.

"There has always been a cultural bias towards knowledge hoarding,"

Charney says. "In the past, businesses and employees have been

territorial about knowledge. It was a competition thing."

Maintain quality assurance. Before implementing the "smarter"

software, you need to establish the integrity of the information already

in place. "The effectiveness of the system depends on the ongoing

quality assurance process," Charney says.

Find out what your customers want. Do your customers want

better product, technical or sales support? Do they want it over the

telephone or on the web? Knowledge tools and software can be deployed

on an enterprise-wide level, helping both employees and customers

in a variety of ways.

Charney suspects the neural network could radically improve

search engines on the web. At this point, however, it is not widely

used outside of private enterprise. It may not be HAL, but then again

— we have not yet reached the millennium.

— Melinda Sherwood

Previous Story Next Story

Corrections or additions?

This page is published by

— the web site for U.S. 1 Newspaper in Princeton, New Jersey.

Facebook Comments