ICT Year 11 Presentation on Expert Systems

A classroom presentation on Expert systems

1. Expert Systems BY : JESSICA Lesson Lesson Lesson

2. Introduction expert system : a computerised system that attempts to reproduce the decision-making process of an expert human Expert systems are made up of five parts: 1 The user interface: an interface that helps the user to communicate with the expert system. It takes the user’s query and sends it to the inference engine. After that, it displays the results. 2 Knowledge base: all of the data, facts and conditions about the expert’s particular subject. The knowledge base is prepared by a knowledge engineer who will be a skilled programmer. 3 A rules base: the set of rules and facts that must be applied to the knowledge base. 4 An inference engine: uses these facts and rules to apply to the knowledge base when selecting an answer to the user’s query. It matches the information that the user inputs at the user interface with data held in the knowledge base to reach an appropriate response by using the inference rules 5 An explanation system: provides reasoning about the information in the knowledge base. It assists the user to understand the contents of the knowledge base and also the reasoning process that produced the solution to the user’s question. It forms a conclusion.

3. Route scheduling for delivery vehicles Medical diagnosis Chess game Financial planning 01 Mineral prospecting Car engine fault diagnosis 02 03 04 05 06 Different scenarios where expert systems are used 07 Financial planning

4. Medical diagnosis 01

5. A medical expert system gives decision support in the form of an accurate diagnosis for the symptoms input. The system receives information (input) about the patient, usually by the patient’s doctor. A medical diagnosis is never a certainty and expert systems have the capability of expressing their conclusions as a probability. It is generally agreed that expert system software must contain a large number of facts and rules about the disease or condition in question in order to deliver accurate answers.

6. Mineral prospecting 02

7. This is when different sites are tested for the probability of minerals, such as ore, being present at a site before mining starts. The expert system can collect its data from: 1 An expert. 2 Directly from sensors that are put in place at the site. 3 Images taken from satellites showing composition of the area such as hills, mountains, lakes and rivers. Data will have to be entered into the system through the user interface. This will include: • The types of rocks that are in the area. • The types of minerals present in the area. • The type of mineral required Using the information, the GIS builds a mathematical model which can predict if the target mineral should be present in association with the features that have been entered. This type of model is called a ‘predictive model’ as it is making a prediction

8. Car engine fault diagnosis 03

9. Car have symbols on their dashboards which light up w hen a sensor detects a problem with a component such as in the engine or the brakes. When the driver takes the car to a garage it is plugged into an expert system, which will interrogate the car’s electronic management system to discover the possible cause of the fault. It may provide several suggestions that the mechanic should investigate. The knowledge in these expert systems can be easily updated online by the vehicle manufacturer, saving time and ensuring it is up to date. It also means that the mechanics do not necessarily need all of the knowledge in the expert system so saving money and time on advanced training courses.

10. Chess game 04

11. A chess player can play a game of chess against the computer; this really means that they can play against an expert system. This can help chess players train without the need for another person - there will be many levels from easy to very difficult – for the player to choose from. The system will have been made by putting information from many expert chess players into the knowledge base.

12. Financial planning 05

13. Expert systems can be very useful for helping individuals and large companies with financial planning. Information about financial situations and personal circumstances is entered via the user interface, and based on each user’s answers, the expert system can suggest a financial plan and give advice on how best they can organise their investments, manage their debts or pay less taxes.

14. Route scheduling for delivery vehicles 06

15. This type of expert system would calculate the most efficient route for a delivery vehicle, taking into consideration the areas for delivery. The user interface would need to ask for: • The amount of individual deliveries. • The distance between each delivery drop-off point. • Details of the delivery vehicle’s type and speed. • The amount of time available altogether. • Layout of the area such as if it is hilly, so that if the system is the type that can also find the most economical route by calculating fuel economy, it can produce the choice of options between fastest route and most economical route.

16. Plant and animal identification 07

17. These expert systems can be used by individuals, scientists and farmers to identify plants and animals. They could help them identify harmful insects and weeds and help to prevent crop failures. If there is a plant or an animal that isn’t recognised, an expert system will ask the user , questions about the animal or plant eg.in the case of an animal, how many legs it has, how long its legs are, does it eat meat or vegetation, its approximate size and colour, etc.

18. THANKYOU ! ! ! !