Know and understand characteristics, uses and purpose of expert systems including mineral prospecting, car engine fault diagnosis, medical diagnosis, chess games, financial planning, route scheduling for delivery vehicles, plant and animal identifica

6 ICT Applications – Expert Systems

Learning Objective

Know and understand the characteristics, uses and purpose of expert systems, with particular reference to the following applications:

  • Mineral prospecting

  • Car engine fault diagnosis

  • Medical diagnosis

  • Chess games

  • Financial planning

  • Route scheduling for delivery vehicles

  • Plant and animal identification

Purpose of Expert Systems

Expert systems provide decision‑support in domains where human expertise is scarce, costly or time‑critical. They enable faster, more consistent and transparent problem‑solving by capturing specialist knowledge in a reusable computer program.

What Is an Expert System?

An expert system is a computer program that mimics the decision‑making ability of a human expert. It combines a specialised knowledge base with a reasoning engine that can draw conclusions from that knowledge.

Core Components

  • Knowledge base – factual information about the domain (e.g., geological formations, medical symptoms).

  • Rules base – a collection of if‑then rules or case‑based examples that encode expert reasoning.

  • Inference engine – applies logical reasoning to the knowledge and rules bases. It can operate by forward‑chaining (data‑driven) or backward‑chaining (goal‑driven).

  • Explanation system – generates a human‑readable justification for each conclusion, helping users understand the reasoning process.

  • User interface – the front‑end through which users input data and receive advice or solutions.

Key Characteristics

  • Domain‑specific knowledge captured from human experts.

  • Rule‑based or case‑based reasoning (if‑then statements, similarity matching).

  • Transparent decision‑making – the system can explain how a conclusion was reached.

  • Consistent performance – no fatigue, emotion or bias.

  • Up‑datable – new knowledge can be added as the domain evolves.

  • Subject to the knowledge‑acquisition bottleneck: obtaining, verifying and maintaining high‑quality knowledge is often the most difficult part of development.

Reasoning Methods

  • Forward‑chaining – data‑driven; starts with known facts and applies rules to generate new facts until a goal is reached. Common in diagnostic systems.

  • Backward‑chaining – goal‑driven; works backwards from a hypothesis, looking for data that can support it. Frequently used in troubleshooting.

  • Hybrid systems can combine both methods for greater flexibility.

Common Purposes (Why Use an Expert System?)

  • Assist professionals in complex decision making.

  • Provide expert advice where human experts are scarce or expensive.

  • Increase speed and accuracy of problem solving.

  • Reduce costs associated with trial‑and‑error approaches.

Applications of Expert Systems

ApplicationDomainHow the System WorksKey Benefits
Mineral ProspectingGeology / MiningAnalyses geological data (rock types, seismic readings, chemical assays) using a rule‑based model to predict high‑potential mineral zones.Reduces costly exploratory drilling; focuses resources on the most promising sites.
Car Engine Fault DiagnosisAutomotive EngineeringCollects sensor readings and driver‑reported symptoms, matches them against a fault‑rules base, and suggests probable causes and repairs (often via backward‑chaining).Speeds up repair time; supports less‑experienced technicians.
Medical DiagnosisHealthcareGathers patient symptoms, test results and history; applies diagnostic rules (forward‑chaining) to generate a ranked list of possible conditions and recommended investigations.Improves diagnostic accuracy; offers decision support for rare or complex cases.
Chess GamesArtificial Intelligence / GamingUses opening libraries, end‑game tablebases and evaluation functions; searches possible move trees to select the optimal move.Provides a strong opponent for learners; demonstrates strategic thinking and AI techniques.
Financial PlanningFinance / Personal AdvisoryInputs client goals, income, expenses and risk tolerance; applies financial‑rules to produce investment, savings and retirement plans.Delivers personalised advice; helps users make informed financial decisions.
Route Scheduling for Delivery VehiclesLogistics / TransportProcesses delivery addresses, vehicle capacities and real‑time traffic data; uses optimisation algorithms (e.g., travelling‑salesman, vehicle‑routing) to generate efficient routes.Reduces fuel costs; improves delivery timeliness and customer satisfaction.
Plant and Animal IdentificationBiology / EcologyAsks a series of characteristic questions (leaf shape, habitat, colour, etc.) and matches answers to a species database using case‑based reasoning.Assists students, researchers and the public in accurate species identification.

Example Reasoning Process – Medical Diagnosis (Forward‑Chaining)

  1. Collect patient data – symptoms, vital signs, test results.

  2. Match each datum against the rule set (e.g., If fever ≥ 38 °C and cough = true then consider respiratory infection).

  3. Accumulate supporting evidence for each possible condition.

  4. Rank diagnoses by confidence level.

  5. Present the most likely diagnoses together with an explanation of the rules applied.

Advantages and Disadvantages (AO3 Evaluation)

AdvantagesDisadvantages

  • Consistent, unbiased decisions

  • Fast processing – 24/7 availability

  • Can handle large knowledge bases

  • Transparent reasoning via explanation system

  • Quality depends on the knowledge base – the knowledge‑acquisition bottleneck

  • Difficulty with ambiguous, incomplete or novel situations

  • Limited creativity compared with human experts

  • Maintenance and updating can be costly

Evaluating an Expert‑System Solution (Guiding Questions)

  • How does the system improve decision speed and accuracy compared with a human expert?

  • What cost savings are realised, and are there any hidden costs (e.g., knowledge‑base maintenance)?

  • To what extent does the explanation system increase user confidence?

  • Are there risks associated with over‑reliance on the system (e.g., missed rare conditions in medical diagnosis)?

  • How does the knowledge‑acquisition bottleneck affect long‑term viability?

Suggested Diagram

Block diagram of an expert system showing the Knowledge Base, Rules Base, Inference Engine, Explanation System and User Interface, with data‑flow arrows between them.

Review Questions

  1. List and describe the five core components of an expert system.

  2. Explain the difference between forward‑chaining and backward‑chaining reasoning, giving an example of an application that uses each method.

  3. Discuss two ways an expert system can improve mineral prospecting, and identify one limitation it may face.

  4. Identify two benefits and two drawbacks of using expert systems for medical diagnosis.

  5. Describe how a route‑scheduling expert system uses optimisation techniques to generate efficient delivery routes.

  6. Give an example of how a plant‑identification expert system interacts with the user to reach a conclusion.

  7. Evaluate the overall impact of expert systems on a chosen application (e.g., financial planning vs. chess playing), weighing both advantages and limitations.