Expert system on diagnosing children’s illness using Bayesian Method

Abstract

Expert system is a computer program that replicates the process of thinking and expert knowledge in solving a particular problem. The expert system can be implemented in wide range of areas, including in the medical field. Children are very susceptible to germs and parents often fear of not being able to detect symptoms of illness usually suffered by children. This expert application system for paediatric diagnosis is an expert system application designed to assist parents to sign in as users and enable them to diagnose the type of illness suffered by children according to the perceived symptoms based on a dynamic knowledge. This knowledge is obtained from various sources including scientific journals, the results of interviews and intensive discussions with medical experts. The knowledge base is structured systematically into a database with several tables including disease tables, symptom tables, and tables on basic rules to support system performance in drawing conclusions. The conclusion in this expert system is drawn using Bayesian Method. The assessment on 30 cases shows that the accuracy of system output is 76.3 percent and the system evaluation on users is 85 percent.