Knowledge engineering has created systems that do everything from focusing
the lens on a camcorder to helping blind people use the Internet. From robotics
to satellites to business computers, knowledge engineers will make existing
technology more useful in the future.
Knowledge engineers use tools like fuzzy logic, neural networks, genetic
algorithms, online diagnostics and simulation programs. They collect information
from human experts and professionals. They need to have an understanding of
the tools they are using, the systems they are designing and the results they
are trying to achieve.
Fuzzy logic is the backbone of this career. It assigns a degree of truth
to different factors. So something can be 80 percent true or 10 percent true.
Knowledge engineer Richard Poppen says computers typically use mathematical
logic: something is either 100 percent true or 100 percent false. But most
humans don't think like that, and fuzzy logic gives computers the knowledge
base they need to solve those problems and make "human" choices.
For example, the sentence "John is tall" is not clear; it's a comment people
take for granted. For a computer to understand "tall," you need to program
it with something like:
IF John is 6'5", then John is tall. IF John is 5'4", then John is not
tall.
Any height in between can be given a percentage rating of tall. For example,
at a height of 6'2", the computer might give a 96 percent certainty that John
is tall. "Fuzzy logic makes rules for making inferences," says Poppen.
When a computer is given this sort of artificial intelligence, it can be
programmed to see and recognize things and respond to spoken commands. Link
more than one computer and knowledge base together, and you've created a neural
network with even more potential for artificial intelligence.
Knowledge engineering is used in the field of environmental management
and forestry. Using software models to simulate the effects of environmental
change on forests, a plan can be created to preserve balance in that specific
ecosystem. Programs can be designed to recognize different kinds of fungi
based on their appearance.
Another place that knowledge engineering is used is in the field of sales
and marketing. Dermot Bradley works for a firm specializing in helping businesses
increase sales. He created a model that helps sales teams navigate the politics
involved in big business.
His rule-based system captures the expertise and knowledge of planning
large sales by modeling the complicated relationships between products, customer
needs and people in both the buying and the selling organization.
The relationships that the system analyzes include who can give access
to the key decision-makers, what aspects of the decision they can influence
and who may want to deny that access. The system "thinks" about what should
be done to manage potential conflicts and to meet the needs of the potential
customer.
Artificial intelligence and robotics are other fields that employ knowledge
engineers. The Jetsons' maid robot is the first thing that comes to many people's
minds when they think robots, but this field involves much more than that.
Knowledge engineers work in private industry, universities and research
centers and as independent consultants.
Knowledge engineers and other systems analysts typically work a 40-hour
week in an office setting. However, given the technology available today,
more work can be done from remote locations using modems, laptops, electronic
mail and the Internet.
Sometimes extra hours are necessary. "When you're getting up to some sort
of deadline, you can work 60 to 70 hours a week. If you don't enjoy doing
that, this isn't the job for you," says Linnea Dunn, a knowledge engineer
with a nonprofit organization based in California.
Because knowledge engineers spend long periods of time in front of a computer
terminal typing on a keyboard, they're susceptible to eyestrain, back discomfort
and hand and wrist problems -- the same complaints other frequent computer
users have.