Other Software : Expert Systems  

An expert system is an artificially intelligent computer system that stores and can apply knowledge about a particular area of human expertise. Expert systems store data collected from human experts in a knowledge base. An inference engine is used to apply this knowledge to solve problems posed by the user. Most expert systems ask the user questions and then suggest a conclusion to the user, based on the answers given to these questions. An expert system may be able to learn from experience, developing new knowledge or modifying existing knowledge based upon how accurate its conclusions have proved to be in the past.

The knowledge base is generally composed of rules that form conclusions based on the inputs made by the user. A simple example rule might be :

IF the person is aged 60 or over AND is female THEN she is entitled to a pension

An expert system will usually require many inputs and apply many rules before offering a possible conclusion to a user. The expert system may change the questions that it asks the user, based on the answers entered to previous questions. Here are some example uses that expert systems can be put to :

Example Description
Medical Diagnosis The task of identifying the cause of a patient's symptoms is normally carried out by a doctor. However in some areas and countries there is a shortage of doctors, and although doctors have good general diagnostic skills it is not possible for each doctor to be an expert in every field of medicine. An expert system can be used to help diagnose a patient's illness by asking the patient questions about the symptoms. A knowledge base can then be used to identify likely causes. The expert system can benefit from the knowledge of many doctors. However, unlike a real doctor it can not see the patient's symptoms for itself. Instead it relies on the patient to describe their own symptoms.
Customer Technical Support If you experience a problem with a computer system you will probably need to get support from the company that made it. Usually this is done by telephoning a central call centre. The staff at the call centre will have some technical knowledge but it is unlikley that they will have a detailed knowledge of every computer that the company has ever made. Therefore they will probably use an expert system to help answer enquiries. The expert system will be able to store detailed knowedge about every product and add to its knowledge base by learning what solutions have worked for particular problems in the past. This will enable members of staff who don't know very much about a particular product to help users solve problems with it. The member of staff will ask the customer questions, prompted by the expert system and the system will attempt to diagnose likely causes and appropriate courses of action based on the customer's answers. The expert system will help the company serve its customers better but if a problem occurs which has never been encountered before a human expert may need to investigate the cause.

Expert systems are useful because

However an expert system is only as good as the knowledge it contains, and expert systems may not recognise unusual situations that a human expert would spot.

GCSE ICT Companion 04 - (C) P Meakin 2004