Expert

Systems

Knowledge Based Systems

Expert Systems
Chapter 6
 Expert
Systems
Content
Knowledge Based Systems

What is an Expert System?

Characteristics of an Expert System.
Classification of Expert Systems.
Components of an Expert System.
Advantages & Disadvantages of Expert Systems.
Creating an Expert System.
 Expert
Systems
Content
Knowledge Based Systems

What is an Expert System?

Characteristics of an Expert System.
Classification of Expert Systems.
Components of an Expert System.
Advantages & Disadvantages
Creating an Expert System.
 Expert
Systems
Expert System
Knowledge Based Systems

Computer software that:

 Emulates human expert
 Behave as same as human

 Deals with small, well defined domains of

expertise
 Medicine, engineering, production

 Is able to solve real-world problems

 Expert
Systems
Expert System
Knowledge Based Systems

 Is able to act as a cost-effective consultant

 No need to be hired consultant

 Can explains reasoning behind any solutions

it finds
 Solutions with reasons

 Should be able to learn from experience.

 Should learn from previous actions &

solutions
 Expert
Systems
Expert System
Knowledge Based Systems

An expert system is a system that employs

human knowledge captured in a computer to
solve problems that ordinarily require
human expertise.(Turban)
 Expert
Systems
Expert System
Knowledge Based Systems

A computer program that emulates the

behaviour of human experts who are solving
real-world problems associated with a
particular domain of knowledge.
(Pigford & Braur)
 Expert
Systems
Who is an expert?
Knowledge Based Systems

Who is an expert?

“An expert is someone who has a special skill

or special knowledge of subject, gained as a
result of training and experience”
 Expert
Systems
What is an Expert?
Knowledge Based Systems

Solve simple problems easily.

Use their knowledge efficiently
Explain why conclusion reached.
Talk easily with other experts in their field.
Learn from experience.
Transfer knowledge from one domain to another.
 Expert
Systems
Expert System
Knowledge Based Systems

Expert Systems manipulate knowledge while

conventional programs manipulate data.
 Expert
Systems
Content
Knowledge Based Systems

What is an Expert System?

Characteristics of an Expert System.
Classification of Expert Systems.
Components of an Expert System.
Advantages & Disadvantages
Creating an Expert System.
 Expert
Systems
Knowledge and Uncertainty
Knowledge Based Systems

Facts and rules are structured into a knowledge

base and used by expert systems to draw
conclusions.
There is often a degree of uncertainty in the
knowledge.
 Things are not always true or false
 the knowledge may not be complete.

Therefore, expert systems can not make

conclusions if its not structured in facts & rules.
 Expert
Systems
Content
Knowledge Based Systems

What is an Expert System?

Characteristics of an Expert System.
Classification of Expert Systems.
Components of an Expert System.
Advantages & Disadvantages
Creating an Expert System.
 Expert
Systems
Content
Knowledge Based Systems

What is an Expert System?

Characteristics of an Expert System.
Classification of Expert Systems.
Components of an Expert System.
Advantages & Disadvantages
Creating an Expert System.
 Expert
Systems
Classification of Expert System
Knowledge Based Systems

Classification based on “Expertness” or

Purpose
Expertness
theused
userfor routine
talks analysis
over the
the
anduser
problem accepts
points
with out the
the those
 An assistant system’s
system untiladvice
portions ofa the work
“joint
 A colleague without
decision” question.
where the human
is reached.
 A true expert expertise is required.
 Expert
Systems
Content
Knowledge Based Systems

What is an Expert System?

Characteristics of an Expert System.
Classification of Expert Systems.
Components of an Expert System.
Advantages & Disadvantages
Creating an Expert System.
 Expert
Systems
Content
Knowledge Based Systems

What is an Expert System?

Characteristics of an Expert System.
Classification of Expert Systems.
Components of an Expert System.
Advantages & Disadvantages
Creating an Expert System.
 Expert
Systems
Components of an Expert
Knowledge Based Systems System

Expert System

Knowledge
Base
User
Interface
Inference
Engine

User
 Expert
Systems
Content
Knowledge Based Systems

What is an Expert System?

Characteristics of an Expert System.
Classification of Expert Systems.
Components of an Expert System.
Advantages & Disadvantages
Creating an Expert System.
 Expert
Systems
Content
Knowledge Based Systems

What is an Expert System?

Characteristics of an Expert System.
Classification of Expert Systems.
Components of an Expert System.
Advantages & Disadvantages
Creating an Expert System.
 Expert
Systems
Desirable Features of an
Knowledge Based Systems Expert System
Give explanation
Ease of Modification
Can be modified easily
Transportability
Adaptive learning
 Expert
Systems
Advantages
Knowledge Based Systems

Superior problem solving

Reliability
Work with incomplete information
Transfer of knowledge
 Expert
Systems
Limitations
Knowledge Based Systems

Expertise hard to extract from experts

 don’t know how
 don’t want to tell

Knowledge not always readily available

 Expert
Systems
Limitations (cont…)
Knowledge Based Systems

High development costs

Only work well in narrow domains
Can not learn from experience
Not all problems are suitable
 Expert
Systems
Content
Knowledge Based Systems

What is an Expert System?

Characteristics of an Expert System.
Classification of Expert Systems.
Components of an Expert System.
Advantages & Disadvantages
Creating an Expert System.
 Expert
Systems
Content
Knowledge Based Systems

What is an Expert System?

Characteristics of an Expert System.
Classification of Expert Systems.
Components of an Expert System.
Advantages & Disadvantages
Creating an Expert System.
 Expert
Systems
Creating an Expert System
Knowledge Based Systems

Two steps involved:

1. extracting knowledge and methods from the
expert (knowledge acquisition)
2. Showing knowledge/methods into an
organised form (knowledge representation)
 Expert
Systems
Acquiring the Knowledge
Knowledge Based Systems

What is knowledge?
Data:
 Raw facts, figures, measurements
Information:
 Refinement and use of data to answer specific
question.
Knowledge:
 Refined information
 Expert
Systems
Sources of Knowledge
Knowledge Based Systems

documented
 books, journals, procedures
 films, databases

undocumented
 people’s knowledge and expertise
 people’s minds, other senses
 Expert
Systems
Levels of Knowledge
Knowledge Based Systems

Shallow (not too deep) level:

 very specific to a situation Limited by IF-THEN
type rules.
Deep Knowledge:
 problem solving. Built from a range of inputs
 Based on collected knowledge

.
 Expert
Systems
Good knowledge
Knowledge Based Systems

Knowledge should be:

 accurate
 nonredundant

 consistent

 as complete as possible
(or certainly reliable enough
for conclusions to be drawn)
 Expert
Systems
Knowledge Acquisition
Knowledge Based Systems Difficulties
 Knowledge is not easy to acquire or maintain
 More efficient and faster ways needed to acquire
knowledge.
 Transferring knowledge from one person to
system is difficult
 Expert
Systems
Other Problems
Knowledge Based Systems

Other Reasons
 experts busy or unwilling to part with
knowledge.
 collection should involve several sources not
just one.
 Expert
Systems
Organizing the Knowledge
Knowledge Based Systems

Representing the knowledge

 Rules
 Semantic Networks

 Frames

 Propositional and Predicate Logic

 Expert
Systems
Representing the Knowledge
Knowledge Based Systems

Rules

If
pulse is absent and breathing is absent
Then
person is dead.
 Expert
Systems
Representing the Knowledge
Knowledge Based Systems

Semantic Networks

Owns
Sam Car

Is a

Honda
Colour
Made in
Green
Japan
 Expert
Systems
Representing the Knowledge
Knowledge Based Systems

Frames
based on objects
objects are arranged in a hierarchical manner

Frame Name Vacation

Where Albury
When March
Cost $1000
 Expert
Systems
Representing the Knowledge
Knowledge Based Systems

Propositional & Predicate Logic

based on calculus

J = Passed assignment
K = Passed exam
Z = J and K
Student has passed assignment and passes
exam
 Expert
Systems

Knowledge Based Systems

THAT IS ALL FOR TODAY…

ANY QUESTION?