IGCSE Computer Science 0478 — 6.3 Artificial Intelligence Hangman

The 24 key terms

1artificial intelligence

A branch of computer science dealing with the simulation of intelligent behaviour by computers. The computer imitates intelligence; it is not alive.

2intelligent behaviour

A computer does a task that normally needs human thinking, like diagnosing a fault or choosing a game move, but it only simulates thinking and does not truly understand.

3Two types of AI: expert systems and machine learning

The 0478 course covers only two types of AI. An expert system uses stored facts and rules to give advice. Machine learning learns from data to improve. Other kinds of AI are not examined.

4collection of data and rules

An AI system stores data (facts) and rules that say how to use them, such as a medical system matching symptoms to illnesses. Data without rules is not AI.

5reasoning

An AI system reasons when it applies its rules to its data to work out a new conclusion that was not stored directly, instead of just looking up a saved answer.

6learn and adapt

An AI characteristic where a system changes its own behaviour as it gains new data or experience, so it improves over time. Some AI systems can do this, but it is not required.

7Characteristics of AI

The three main characteristics of AI are a collection of data and rules, the ability to reason, and possibly the ability to learn and adapt. Learning and adapting is optional.

8identify AI characteristics

Read a scenario and find the data and rules, the reasoning, and any learning. Then quote evidence from the description, using the word because to link each quote to the characteristic it shows.

9expert system

An AI system that uses the knowledge and rules of human experts to give advice or answers on one specialist subject. It asks the user questions and gives a conclusion.

10specialist knowledge

The expert knowledge that expert systems supply in areas like medical diagnosis, fault finding, mineral prospecting and financial advice, so the user does not need to be an expert.

11knowledge base, rule base, inference engine, interface

The four parts of an expert system: the knowledge base holds facts, the rule base holds rules, the inference engine reasons, and the interface talks to the user.

12knowledge base

The part of an expert system that stores facts and information about a subject, gathered from human experts. It can be updated with new facts. The rule base holds the rules.

13rule base

The part of an expert system holding IF-THEN rules (IF condition THEN conclusion) that link facts to conclusions, often joined with AND/OR, which the inference engine uses to reach answers.

14inference engine

The part of an expert system that reasons: it matches the user's answers and the knowledge base against the rule base to reach a conclusion. It is the 'brain' of the system.

15user interface

The part of an expert system the user sees. It shows questions, lets the user enter answers and facts, and displays the conclusion in a simple way for non-experts.

16interface -> inference engine -> interface

The user answers questions at the interface; the inference engine applies the rules to the answers and knowledge base; the interface then shows the result.

17trace an expert system

Follow the user's answers through the IF-THEN rules in order, checking which rules fire and which do not, until one rule gives the final conclusion.

18machine learning

A type of AI where a program automatically adapts its own processes and/or data, improving with experience, with no programmer changing it.

19adaptive processes

A machine learning program changes its own processes, meaning how it makes decisions, using what happened before, so no human has to rewrite the code.

20self-updating data

A machine learning program changes its own stored data by adding, updating or removing items as it meets new cases, e.g. spam marked by users, and uses that data in later decisions.

23Expert system vs machine learning

An expert system reasons using rules and facts put in by experts. Machine learning adapts its own processes and data by itself. Both are types of AI.

24classify AI system

Decide if a system gives advice from a knowledge base and rule base (expert system) or changes its own processes or data (machine learning), then justify with because.

25AI system operation

Naming the parts of an AI system, what each does, and how they work together to take input, reason on data and rules or learn, and output a result that simulates intelligence.

26key term + reason

Use exact key terms (data, rules, reasoning, knowledge base, rule base, inference engine, interface, machine learning) and link each point to a reason.