Know and understand components of an expert system: user interface, inference engine, knowledge base, rules base, explanation system

6 ICT Applications – Expert Systems

Learning Objective (AO1)

Know and understand the components of an expert system: user interface, inference engine, knowledge base, rules base, and explanation system. In addition, be able to describe knowledge‑acquisition methods, limitations, and links to safety & security, emerging technologies and societal impact (AO3).

What is an Expert System? (AO1)

An expert system is a computer program that mimics the decision‑making ability of a human expert. It captures specialised knowledge in a particular domain and applies logical reasoning to solve problems, give advice, or make predictions.

Curriculum‑approved Examples (AO1)

• Medical diagnosis system – e.g., a system that asks for symptoms and suggests possible diseases such as malaria or influenza.
• Plant‑identification system – e.g., a system that asks for leaf shape, height and habitat to identify a species of tropical plant.

Key Components (AO1)

1. User Interface (UI)

• Collects input (questions, data) and displays output (answers, recommendations).
• Features: input validation, help prompts, result formatting, colour‑coding for warnings.
• Design goal: make the system usable by non‑experts.

2. Inference Engine

• Forward‑chaining – data‑driven; starts with known facts and fires all applicable rules until a conclusion is reached.
• Backward‑chaining – goal‑driven; starts with a hypothesis and works backwards to find supporting facts.
• Rule‑conflict resolution – when several rules could fire, the engine selects one using strategies such as:

  • Specificity (more specific rule wins)
  • Recency (most recently added rule wins)
  • Priority/weight assigned by the knowledge engineer

3. Knowledge Base

Contains the factual information of the domain. Knowledge is represented in a structured way to enable efficient retrieval.

Knowledge Representation

• Facts – atomic pieces of information (e.g., leafShape = ovate).
• Objects – entities such as Plant or Patient.
• Attributes – properties of objects (e.g., height, temperature).
• Relationships – links between objects (e.g., Plant → hasHabitat → Tropical).
• Often stored in tables, relational databases or ontologies.

4. Rules Base

• Encoded as IF‑THEN statements that capture expert reasoning.
• May include:

  • Certainty factors (e.g., 0.8 confidence)
  • Priority levels for conflict resolution

• Organised by functional sections (diagnosis, treatment, classification, etc.).

5. Explanation System

• Produces a step‑by‑step trace of which rules fired, the facts that triggered them and the order of execution.
• Answers “why?” queries – e.g., “Why was malaria diagnosed?”
• Provides an audit trail useful for debugging, verification and meeting data‑protection requirements.

Knowledge Acquisition (AO2)

Before building the system, expert knowledge must be captured.

• Interviews and structured questionnaires with domain experts.
• Observation of expert work processes (shadowing, video analysis).
• Use of knowledge‑engineering tools (ontology editors, rule‑authoring environments).

Maintenance Cycle (AO3)

1. Monitor – detect changes in the domain (new diseases, new plant species).
2. Update – modify facts in the knowledge base and edit/add rules.
3. Version control – keep records of each update for traceability.
4. Test – re‑run test cases to verify that accuracy is maintained.
5. Deploy – release the updated system to users.

Limitations & Risks (AO3)

• Knowledge‑acquisition bottleneck – extracting tacit expertise can be time‑consuming and costly.
• Maintenance costs – the knowledge base must be regularly updated as the domain evolves.
• Data‑quality dependence – inaccurate input data leads to wrong conclusions.
• Security & privacy concerns – the knowledge base may contain sensitive facts.

Security Controls (AO3)

• Encryption of the knowledge base (at rest and in transit).
• Role‑based access control – only authorised users can view or edit rules/facts.
• Strong authentication (passwords, two‑factor).
• Audit logging of all changes to the rules base and knowledge base.
• Regular backups and integrity checks.

Connections to Other Syllabus Topics (AO3)

• Safety & security / Data protection – encryption, access control and audit trails protect confidential knowledge.
• Emerging technologies (AI & Machine Learning) – expert systems are a symbolic‑AI approach; they contrast with data‑driven machine‑learning models.
• Impact of ICT on society – expert systems can improve decision‑making in medicine, agriculture and engineering, but may reduce the demand for human experts and raise ethical questions.

How the Components Work Together

1. User enters a problem or query via the User Interface.
2. The Inference Engine retrieves relevant facts from the Knowledge Base and applies matching rules from the Rules Base (using forward or backward chaining and conflict‑resolution strategies).
3. Deductions are made, producing a solution, recommendation or prediction.
4. The Explanation System builds an audit‑trail that shows which rules fired and why.
5. The final answer and its explanation are displayed to the user through the UI.

Evaluation of Expert Systems (AO3)

When assessing an expert system, consider the following criteria:

Suggested Diagram

Key Points to Remember (AO1)

• Expert systems separate knowledge (knowledge base) from processing (inference engine).
• Rules are expressed as IF‑THEN statements; they guide the reasoning process.
• The explanation system provides an audit trail, enhancing user confidence and meeting data‑protection requirements.
• A well‑designed user interface makes the system accessible to non‑experts.
• Knowledge acquisition and regular maintenance are critical for reliability.
• Consider safety, security, and societal impact when designing or evaluating expert systems.