Module 12 - Expert Systems

1. Overview of Expert Systems

Expert systems are AI- driven software that use domain- specific knowledge and inference
techniques to solve complex problems typically requiring human expertise. They simulate
the decision- making abilities of a human expert and are widely used across fields like
medicine, engineering, and finance.

1. Representing and Using Domain Knowledge

Domain knowledge refers to specialized information that an expert system relies on to
solve problems. The accuracy, reliability, and efficiency of an expert system largely
depend on how well the domain knowledge is represented and utilized.
Types of Knowledge Representations:
1. Rules ( If- Then Statements) :
These are condition- action pairs that specify what actions to take under certain
conditions.
Example:
In a medical expert system, a rule might be:
If the patient has a sore throat and fever, then suspect a throat infection.
2. Frames ( Object- Oriented Structure) :
Frames use data structures that group information related to a specific object or
concept. They allow easy access to hierarchical information.
Example:
A frame for a patient object might include fields for name, age, symptoms, and
diagnosis. The system can refer to this frame when evaluating similar cases.
3. Semantic Networks:
These represent knowledge in the form of a network with nodes ( concepts) and
edges ( relationships between concepts) . Useful for representing relationships and
associations.
Example:
A financial expert system might have a semantic network where nodes represent
different investment options, and edges represent risk levels or expected returns.
4. Production Rules:
Production rules consist of a set of instructions for actions based on conditions.
They form the basis of a decision tree and help the expert system reach
conclusions.
Example:
In an environmental control expert system, a rule could be:
 IF the temperature exceeds 30°C, THEN activate the cooling system.
Using Domain Knowledge in Expert Systems:
● Inference Engines evaluate the knowledge base and apply logical rules to solve
problems.
● Backward Chaining: Starts with a hypothesis and works backward to confirm it.
● Forward Chaining: Begins with data and applies rules to arrive at a conclusion.
Real- life Example:
A loan approval expert system in a bank uses customer data ( income, credit score,
etc.) and applies rules based on loan eligibility. For instance:
● If credit score > 700 and income > $50, 000, approve loan.

2. Expert System Shells

An expert system shell is a software platform designed to simplify the development of
expert systems by providing a general framework for inference engines and knowledge
representation. Developers use expert system shells to build specific applications by
populating them with relevant domain knowledge.
Key Components of Expert System Shells:
1. Knowledge Base: The knowledge base represents facts and rules. It consists of
knowledge in a particular domain as well as rules to solve a problem,
procedures and intrinsic data relevant to the domain.
2. Inference Engine: The function of the inference engine is to fetch the relevant
knowledge from the knowledge base, interpret it and to find a solution relevant
to the user’ s problem. The inference engine acquires the rules from its
knowledge base and applies them to the known facts to infer new facts.
Inference engines can also include an explanation and debugging abilities.
3. Knowledge Acquisition and Learning Module: The function of this component is
to allow the expert system to acquire more and more knowledge from various
sources and store it in the knowledge base.
4. User Interface: This module makes it possible for a non- expert user to interact
with the expert system and find a solution to the problem.
5. Explanation Module: This module helps the expert system to give the user an
explanation about how the expert system reached a particular conclusion.
How Expert Systems Work?
Expert systems operate by following a structured approach:
1. Input Data: Users provide data or queries related to a specific problem or
scenario.
2. Processing: The inference engine processes the input data using the rules in the
knowledge base to generate conclusions or recommendations.
3. Output: The system presents the results or solutions to the user through the
user interface.
4. Explanation: If applicable, the system explains how the conclusions were reached,
providing insights into the reasoning process.

Reasoning Strategies used by Inference Engine

Forward Chaining and Backward Chaining, which are two fundamental methods for
processing information and solving problems in an expert system:
1. Forward Chaining
This is a data- driven reasoning approach where the system starts with the available facts
and applies rules to infer new facts or conclusions. It’ s typically used to predict
outcomes or determine what will happen next. An example given is predicting stock
market movements.

2. Backward Chaining
This is a goal- driven reasoning approach where the system starts with a hypothesis or a
goal ( something to prove) and works backward to determine which facts or conditions
would support that conclusion. It’ s often used to diagnose issues by determining the
cause of an observed effect. The examples provided include diagnosing medical
conditions like stomach pain, blood cancer, or dengue.
Popular Expert System Shells:
● CLIPS ( C Language Integrated Production System) : Commonly used in academia
and research for building expert systems.
● Prolog: A programming language with built- in capabilities for building rule-
based expert systems.
● Jess ( Java Expert System Shell) : A rule- based system for the Java platform,
often used in business applications.
Real- life Example:
A company could use an expert system shell to create a hiring advisor that processes
candidate information and applies HR rules to suggest the best candidates. The HR
manager enters details, and the system evaluates them against predefined rules like
required experience and education level.

3. Knowledge Acquisition
Knowledge acquisition is the process of collecting and organizing information from
experts to create the knowledge base of an expert system. This step is crucial, as the
quality of knowledge directly impacts the system' s performance and accuracy.
Techniques for Knowledge Acquisition:
1. Interviews with Experts: Conducting structured interviews to elicit detailed
insights, strategies, and problem- solving methods.
Example:
In building a medical diagnosis expert system, developers might interview
experienced doctors to gather knowledge on symptoms and treatment protocols.
2. Observation of Expert Actions: Watching experts solve problems in real- time and
 noting their actions, decision- making, and reasoning processes.
Example:
For an industrial machinery expert system, engineers observe maintenance
technicians diagnosing issues in complex machines and apply similar logic in the
system.
3. Case- Based Reasoning: Using past cases and outcomes to build rules. This
approach is useful in areas where historical data is readily available.
Example:
A legal expert system for contract disputes might use previous court cases to
make recommendations, basing its logic on similar precedent cases.
4. Documentation and Manuals: Using technical manuals, textbooks, or other written
sources to extract relevant rules and procedures.
Example:
In an expert system for vehicle repair, knowledge acquisition might involve
gathering information from automobile repair manuals and manufacturer
specifications.
5. Machine Learning Approaches: Using AI algorithms to learn patterns from large
datasets and generate rules. This is particularly useful when explicit expert
knowledge is difficult to define.
Example:
In an e- commerce recommendation system, data from customer purchases and
browsing behavior can be used to make personalized product suggestions.
Challenges in Knowledge Acquisition:
● Tacit Knowledge: Experts may find it difficult to articulate all the knowledge
they apply intuitively.
● Knowledge Transfer: Translating the expert' s subjective experiences into clear,
programmable rules can be challenging.
● Knowledge Updates: As fields evolve, it is necessary to periodically update the
knowledge base to maintain the expert system’ s relevance.

Real- life Example of Knowledge Acquisition Process:

In developing a veterinary expert system for diagnosing pet diseases, knowledge engineers
might:
● Interview veterinarians to understand symptoms and treatment options.
● Gather information from veterinary textbooks and journals.
● Use past cases ( with consent) to build a library of diseases, symptoms, and
outcomes.
By applying these methods, the resulting system could assist pet owners or vet
technicians in identifying common pet health issues based on symptoms inputted into the
system.