Experts Systems

An Introduction to Expert Systems

Expert Systems is an interactive and reliable computer-based decision making system which uses both facts and heuristics to solve complex problems.The purpose of an expert system is solve the most complex issues in a specific domain. It is considered at the highest level of human intelligence and expertise.

It's based on knowledge acquired from an expert and capable of expressing and reasoning about some domain of knowledge. It can resolve many issues generally would require a human expert.

Characteristics of Expert Systems

• Highly responsive - An expert system interact with the user in a very reasonable period of time. It must get the most accurate solution for the same problem faster than by the expert.

• High performance - The expert system offers the highest level of expertise. It provides efficiency, accuracy and imaginative problem solving.

• Capable of handling challenging decision & problems - Expert systems are capable of handling challenging decision problems and delivering solutions.

• Good reliability - The expert systems must not make a mistake.

Components of the expert system

The components of expert systems are listed as below,

1. Knowledge Base

2. Inference Engine

3. User Interface

Knowledge Base

It contains all the knowledge about the specific problem domain. It is like a container of knowledge which is obtained from different experts of a specific problem domain. The success of the expert system mainly depends on the highly accurate and precise knowledge.

Key terms of Expert Systems

Facts - A small portion of important information.

Rules - How the facts are selected and applied to a user problem

Components of Knowledge base

The Knowledge base contains two types of knowledge

1. Factual Knowledge - Information widely accepted by the Knowledge Engineer and scholars in a specific problem domain.

2. Heuristic Knowledge - About the practice, judgement, accuracy, guessing capability, and ability of evaluation.

Knowledge representation

The method used to organize and formalize the knowledge in knowledge base. Those are in the form of IF-THEN-ELSE rules.

Knowledge Acquisition

Knowledge acquisition is the process of extracting, structuring and organizing knowledge from a human expert, converting the acquired knowledge into rules (IF-THEN-ELSE) and injecting them into knowledge base.

Participants of Expert systems

Domain Expert - Person or group of people whose expertise and knowledge is taken to develop an expert system.

Knowledge Engineer - Person who integrates knowledge into the system. Also monitors the development process of the expert system.

End User - Person or group of people who are using the expert system at the end. Expert system will provide advice by analyzing the facts without help from experts.

Inference Engine

Inference Engine known as the brain of the expert system. It decides "what can happen next." Inference engine contains rules to solve a specific problem. It integrates with the knowledge base and refers the knowledge from the knowledge base. Then it selects facts and related rules to apply when user gives an query to process. Then it helps to deduct the problem to find the solution and provides reasoning, conclusion about the information in the knowledge base.

User Interface

User Interface provides interaction between the expert system and the user. It takes the user's query in a readable form and passes it to the inference engine. Generally uses Natural Language Processing for that. Then it gives the results to the user. Simply user interface helps the user to communicate with the expert system.

How to build an Expert Systems

• Identify the problem domain. (Characteristics of problem domain)

• Knowledge engineer and domain expert work together to define the problem.

• Knowledge engineer translates the collected knowledge into a computer understandable language (for example, CLIPS)

• Knowledge engineer design the inference engine, the brain of the expert system which can use knowledge when needed.

• Knowledge engineer the test a prototype of expert system with sample cases and also end user will test the prototype.

• Train the end user to use expert system.

• Maintain the expert system up-to-date, by regular review and update on the knowledge base.

Benefits of Expert Systems

• Ability to solve complex and challenging issues.

• Offers fast and accurate solutions.

• Error rate is low when compared to a human.

• Maintains a significant level of information that will really challenging for a human expert to work with.

• Offers consistent solutions for repetitive problems.

• Production cost is reasonable.

• Works steadily without getting emotional, tense, or fatigued.

• Have a proper explanation of decision making.

• Can work in the environment dangerous to humans.

• Cuts the expenses of consulting experts for problem solving.

Limitations of Expert Systems

• Error in the knowledge base can lead to wrong decision.

• Maintenance cost if too expensive.

• Limited for a specific problem domain.

• Unable to make creative responses in an extraordinary situation.

• Very difficult to maintain as the changes occurs in problem domain.

Applications of Expert Systems

• Hospitals and medical facilities.

• Loan analysis.

• User navigation system. (indoor/outdoor)

• virus detection

• Process monitoring control.

• Planning and scheduling.

• Identifying the fault products.

• Fraud detection on finance.

• Stock market trading.