Detailed · Deep study
1Components
| Component | Role |
|---|---|
| user interface | asks the user questions (often yes/no) and shows the results |
| inference engine | applies the rules to the facts and the user's answers to reach conclusions; decides which question to ask next |
| knowledge base | a database of facts about the subject, collected from human experts |
| rules base | the rules (IF … THEN) used to reason with the facts |
| explanation system | explains how the conclusion was reached, often with a probability (e.g. "80% likely") |
2Using an expert system
- The user interface asks a series of questions about the problem, e.g. symptoms.
- The inference engine compares the answers with the knowledge base, using the rules base.
- It chooses the next question based on the answers so far.
- It produces possible solutions or diagnoses, with probabilities.
- The explanation system shows the reasoning.
| Use | What it does |
|---|---|
| mineral prospecting | suggests where to drill or mine from geological data |
| car engine fault diagnosis | identifies faults from sensor readings and symptoms |
| medical diagnosis | suggests possible illnesses and treatments from symptoms |
| chess games | chooses moves using rules and many stored games |
| financial planning | advises on investments, loans and pensions |
| route scheduling for delivery vehicles | plans the most efficient delivery routes |
| plant and animal identification | identifies a species from features the user describes |
| Advantages | Disadvantages |
|---|---|
| consistent answers; does not forget or get tired | no common sense; results only as good as the knowledge and rules |
| available any time; holds the knowledge of many experts | expensive to set up and keep up to date |
| fast; can give a probability for each answer | users may rely on it without checking; no empathy |
3Exam focus
Key words: expert system; user interface; inference engine; knowledge base; rules base; explanation system.
Quick self-check
Answer in your head first, then reveal.
User interface; inference engine; knowledge base; rules base; explanation system (5).
Asks questions / reads sensor data (1); inference engine uses rules base and knowledge base (1); chooses next question (1); gives possible faults with probabilities (1).
