What is Expert System
An expert system (ES) is a computerized information system that allows non-experts to make decisions comparable to that of an expert. Expert systems are used for complex or unstructured tasks that require experience and specialized knowledge.
Components of Expert System
Expert systems typically contain the following components:
- Knowledge base: This includes data, knowledge, relationships, rules of thumb (heuristics) and decision rules used by experts to solve a particular type of problem. A knowledge base in a computer is equal to the knowledge of an expert or group of experts developed through years of experience in their field.
- Inference engine: This program consists of logic and reasoning mechanism that can simulate (=replicate, reproduce) the expert’s logic process and deliver advice. It uses data obtained from both knowledge base and the user to make associations and inferences, form conclusions and recommend a course of action.
- User interface: This program allows the user to design, create, update, use and communicate with the expert system.
- Explanation facility: With the help of this facility user can know the logic being followed by the expert system to arrive at the conclusion.
- Knowledge acquisition facility: Building a knowledge base, known as knowledge engineering, involves both human expert and a knowledge engineer. The knowledge engineer extracts an individual’s expertise and uses the knowledge acquisition facility to enter it into knowledge base.
Advantages of Expert System
- They provide a cost-effective alternative to human experts.
- They can outperform a single expert because their knowledge is gained from several experts. They are faster and more consistent and do not get over worked or stressed out.
- They produce better-quality and more consistent decisions. Expert systems assist users in identifying problems, which increases the probability of making sound decisions.
- They can increase productivity (=i.e. production per unit of time).
- They preserve the expertise of an expert, leaving the organisation.
Disadvantage of Expert System
- Development can be costly and time-consuming. Some large systems require upto 15 years and millions of dollars to develop.
- It is difficult to obtain knowledge from experts because it is very difficult to specify exactly how the decision maker has taken decision.
- Designers were unable to program human being’s common sense into current systems. If the system faces any situation, which is not programmed to handle, then the system may break down.
Example of Expert System
As technology advances, these problems can be overcome and expert systems will play an important role in accounting information systems. Following are examples of companies that have successfully used expert systems:
- IRS analyzes tax returns to determine the returns to be passed to tax fraud investigators.
- IBM designs and evaluates internal controls in both new and existing applications.
- American Express authorizes credit card purchases to minimize fraud and credit losses. Its Expert System has replaced 700 authorization clerks and saved millions of dollars.
