O Level Computer Science (2210)

Automated and emerging technologies

Artificial intelligence

What makes a system intelligent, how expert systems are built, and how machine learning improves from data.

Examples use Cambridge pseudocode for the 2026–2028 syllabus.

What is AI?

Artificial intelligence (AI) is a branch of computer science that deals with simulating intelligent behaviour by computers. An AI system typically has these characteristics:

  • It collects data and uses it.
  • It has rules for using that data.
  • It can reason, applying the rules to make decisions or predictions.
  • It can learn and adapt, changing its own processes and rules based on new data.

Expert systems

An expert system gives advice like a human expert would, for example diagnosing illnesses, identifying minerals, or fixing car faults. It has four parts:

  1. 1
    User interface
    Asks the user questions and shows conclusions.
  2. 2
    Inference engine
    Applies the rules to the user's answers to reach a conclusion.
  3. 3
    Rule base
    IF-THEN rules, such as IF temperature > 38 AND rash THEN ...
  4. 4
    Knowledge base
    Facts collected from human experts.

Exam tip. The inference engine connects the user interface to the rule base and knowledge base. It asks the next question based on earlier answers.

Machine learning

In machine learning, a program is not given every rule. It is trained on data and changes its own processes to get better at a task, finding patterns that a person might not spot.

Examples include recommending films based on what you have watched, filtering spam email, recognising faces in photos, and detecting fraudulent card payments.

  1. 1
    Train on data
    Thousands of labelled examples.
  2. 2
    Predict
    Make a decision on new data.
  3. 3
    Check the result
    Was it right?
  4. 4
    Adapt
    Adjust the model to do better next time.
then back to step 1

Check yourself

Name the four parts of an expert system.Show answer

User interface, inference engine, rule base, and knowledge base.

How does machine learning differ from a normal program?Show answer

A normal program follows fixed rules written by a programmer. A machine learning program changes its own rules based on the data it is trained on.

References used to build these lessons

We used these official documents as references when building this course's lessons. They are the sources for the syllabus structure and course requirements.

Cambridge International · 2210 · Exams 2026–2028

Syllabus reference PDFOfficial source