What are Expert Systems?
Expert systems are developed to imitate the knowledge and expertise of an expert in a particular field. For example:
- tax and financial calculations
- road scheduling for delivery of vehicles
- strategy games (e.g. chess)
- identification of plants, animals or chemical compounds
- prospecting for oil, gas and minerals
- diagnosing a person’s illness
- diagnostics (e.g. finding faults on a circuit board, faults in a car engine, etc)
How to set up an expert system
- Several experts in a given field are first interviewed.
- This results in the collection of data from these experts.
- To store this data, a knowledge base is first designed and created.
- Rules base and interference engines are also designed and created.
- An explanation system is also developed.
- Input and Output screens and formats also known as user interface, are designed and developed.
- The expert system is tested against known conditions and scenarios and is also checked to see whether it meets the original specification.
- Experts are interviewed about the effective of this system before it is officially released.
- They never forget to answer a question when determining the logic.
- They provide consistent results and are not affected by emotional reasoning.
- Expert systems reduce the time to solve a problem.
- The potential of saving money is high since the need of hiring specialists is reduced (e.g., when carrying out oil exploration).
- Allows parts of the world access to expertise which they normally may not be able to afford.
- They indicate the probability of the given solution being acccurate or correct.
- Needs considerable training of the operators to ensure that the system is used correctly.
- The initial setup costs are generally very high.
- They tend to lack common sense in some of the decision-making processes.
- Errors in the knowledge base can lead to incorrect decisions being made.
- Though lack of emotional reasoning is an advantage, it can also be a disadvantage in certain areas like medical diagnosis.
Using an expert system
Taking oil prospecting as a case, the process for using an expert system would be as given below:
- An interactive user screen appears (this is often multiple-choice questions or Yes/No responses).
- The system asks questions about geological profiles.
- The operator keys in the answers to the questions/geological profiles.
- The system then asks questions based on the previous response(s) just keyed in.
- The inference engine compares the answers to questions with the facts or data fed in earlier in the knowledge base using the rules base.
- The system returns with the probability of finding oil as an output.
- It would also indicate the probable depth of the deposits (which is usually as a percentage probability).
- The explanations system would also explain how the expert system arrived with its conclusions.
- It would then make predictions about geological deposits in the soil/rocks.
- Finally, it produces contour maps showing the concentration of minerals, rocks, oil, etc.
*** This is the end of this guide. Hope you enjoyed it! Thank You for using Igcsepro.org. We hope you will give us a chance to serve you again! Thank you! ***