Introduction to

Expert Systems

Almost any inference method will do ; the power resides in

the knowledge.

~ Edward Feigenbaum (Stanford University)

Current Trends in Applied AI

• Expert Systems
• Fuzzy Systems
• Artificial Neural Networks
• Genetic Algorithms and Evolutionary Programming
• Swarm Intelligence Systems
 Soft computing Methods
Fuzzy Sets
Fuzzy Logic

Soft Genetic
Rough Sets Computing
Algorithms

ANN
 Soft Computing
 Soft Computing is a collection of methodologies (working
synergistically, not competitively) which, in one form or another,
reflect its guiding principle:

 Exploit the tolerance for imprecision, uncertainty, approximate

reasoning and partial truth to achieve tractability, robustness, and
close resemblance with human like decision making.

 Provides flexible information processing capability for

representation and evaluation of various real life ambiguous and
uncertain situations. Real World Computing

 It may be argued that it is soft computing rather than hard

computing that should be viewed as the foundation for Artificial
Intelligence (AI).
 Soft Computing vs Hard Computing
 Hard computing requires programs to be written; soft computing can evolve
its own programs

 Hard computing uses two-valued logic; soft computing can use multivalued or
fuzzy logic

 Hard computing is deterministic; soft computing incorporates stochasticity

 Hard computing requires exact input data; soft computing can deal with
ambiguous and noisy data

 Hard computing is strictly sequential; soft computing allows parallel

computations

 Hard computing produces precise answers; soft computing can yield

approximate answers
Expert Systems
“An expert system is a computer system that emulates, or acts in all
respects, with the decision-making capabilities of a human expert.”
Professor Edward Feigenbaum
Stanford University

• Expert systems are built to solve a wide range of problems in

domain such as medicine, math, engineering, chemistry,
geology, computer science, business, low, defense and
education.
Expert Systems

• An expert system differs from other AI systems

in that it attempts to explicitly embody expertise
and knowledge with in software.
Fuzzy Systems

As complexity rises, precise statement lose

meaning and meaningful statement lose precision
~ Lotfi A. Zadeh : Fuzzy Sets (1964)

• Analysis of data can be subjective

• Fuzzy systems allow an element of fuzziness or vagueness to
be associated with data
• Useful application – decision making under uncertain or
hypothetical situations
 Fuzziness in everyday world
 John is tall
 Temperature is hot
 Temperature is cold
 Mr. Ram is young
 My car is slow, but your car is fast
Fuzzy Sets (Lotfi A. Zadeh , 1965)
Fuzzy Sets Conventional uncertainty models

capture vagueness exclusively through membership values.

Artificial Neural Networks (ANN)

• 1943 onwards – McCulloch & Pitts

Model of an elementary computing neuron
• ANN are often employed where a problem
requires estimation, prediction, or classification
• ANN do not store knowledge or expertise
explicitly but implicitly characterize behavior
through a learning process
Artificial Neural Networks
 Artificial Neural Networks

Models of the brain and nervous system

 The brain is a highly complex, nonlinear, and

parallel information-processing system.

 It has the capability of organizing neurons so as to

perform certain computations (e.g. pattern
recognition, perception, and motor control) many
times faster than the fastest digital computer.
Model of a Neuron
Genetic Algorithms and Evolutionary Programming

• 1950s-1960s: Evolutionary systems (optimization tool

for engineering problems)

• GAs are favored for the search problems which require

the identification of a global optimal solution

• GAs are based on natural evolution process and

operate by combining the best solutions at each stage
to obtain the next generation of solutions
Evolutionary Computation
Swarm Intelligence Systems

• Swarm Intelligence utilizes the behavior of non-

human living entities for problem solving -Ant
Colony Optimization (ACO) & Particle Swarm
Optimization (PSO)
 Expert Systems

 Expert systems are computer programs built for

commercial application using the programming

techniques of AI, especially those developed for problem

solving.

 It is called a system because it consists of both a

problem-solving component and a support

component
Expert Systems - Features

• Facility for non-expert personal to solve problems

that require some expertise
• Speedy solutions
• Reliable solutions
• Cost reduction
• Elimination of uncomfortable and monotonous
operations
• Wider access to knowledge
When to Use an Expert System
 21

Problem Domain vs. Knowledge

Domain
• An expert’s knowledge is specific to one problem
domain – medicine, finance, science, engineering, etc.

• The expert’s knowledge about solving specific

problems is called the knowledge domain.

• The problem domain is always a superset of the

knowledge domain.
 22

Representing the Knowledge

The knowledge of an expert system can be

represented in a number of ways, including IF-
THEN rules:

IF you are hungry THEN eat

 23

Knowledge Engineering

The process of building an expert system:

1. The knowledge engineer establishes a dialog with

the human expert to elicit knowledge.

2. The knowledge engineer codes the knowledge

explicitly in the knowledge base.

3. The expert evaluates the expert system and gives a

critique to the knowledge engineer.
 24

Development of an Expert System

 25

Early Expert Systems

• DENDRAL – used in chemical mass spectroscopy to

identify chemical constituents

• MYCIN – medical diagnosis

• DIPMETER – geological data analysis for oil

• PROSPECTOR – geological data analysis for minerals

• XCON/R1 – configuring computer systems

Expert Systems Organization
Expert Systems
Expert Systems - Components

Knowledge base
Stores all relevant information, data, rules, cases, and
relationships used by the expert system
Knowledge acquisition facility
Provides a convenient and efficient means of capturing and
storing all components of the knowledge base
Inference engine
Seeks information and relationships from the knowledge
base and provides answers, predictions, and suggestions in
the way a human expert would
Explanation facility
A part of the expert system that allows a user or
decision maker to understand how the expert system
arrived at certain conclusions or results
Expert Systems - Features

• We can represent the knowledge of a expert

system using rules and objects.

IF-THEN Rule:
IF the load demand is medium, THEN the
system is reliable
Rule can pattern-match on objects as well as
facts
 Expert Systems

The process of building an expert system is called Knowledge

Engineering and is done by knowledge engineer

Expert Systems - Characteristics

• High performance
• Expertise
• Adequate response time
• Good reliability
• Self-knowledge
• Understandable
• Justification
• Flexibility
Expert Systems
Expert Systems Development
 Building a Rule-based Expert System

Knowledge Engineering
• People involved -Knowledge Engineer, Domain Expert
and End User

• Rules:
IF you want to get good marks and a good job
THEN work hard
• Represent the Knowledge (rules or frames)
• Determine the search strategy (backward or forward)
 Building a Rule-based Expert System

The Knowledge base: Core of an Expert System

 Expert System Architecture – production system
 Knowledge base - set of production rules

The inference engine: Recognize-act part of the

production system
 Data-driven or Goal-driven
Verification and Validation of Expert
Systems

Validation: shows the correctness of the final

product with respect to the users needs and
requires to deal with whether the correct
problem was solved or not
Verification: ensuring that an expert system has
been developed correctly and does not contain
technical errors
Validation Vs Verification

Validation means building the right system.

Verification means building the system right.

~ O’Keefee et al.
Applications of Expert Systems and
Artificial Intelligence
Credit granting
Information management and retrieval
AI and expert systems embedded in products
Plant layout
Hospitals and medical facilities
Help desks and assistance
Employee performance evaluation
Loan analysis
Virus detection
Repair and maintenance
Shipping
Marketing
Warehouse optimization
 Limitations of Expert Systems
• Lack of general knowledge
• Not widely used or tested
• Limited to relatively narrow problems
• Cannot readily deal with “mixed” knowledge
• Possibility of error
• Cannot refine own knowledge base
• Difficult to maintain
• May have high development costs
• Raise legal and ethical concerns
References

• Artificial Intelligence and Intelligent Systems, N.

P. Padhy, Oxford University Press, 2005