A Dempster-Shafer Approach for Uncertainty Management in Diagnosing Eye Diseases

Expert system can be used as a consultant that could give an advice to users as well as an assistant to experts. One way to help diagnose eye disease is to create an expert system as a media consultation therefore the eye disease could be identified based on the symptoms. Expert system in this research is built using using forward chaining search method and dempster shafer method to handle the uncertainty. The dempster shafer method is a non monotonic method that is used to look for inconsistencies due to the addition or subtraction of the new facts that will change the existing rules. This research aims to handle the uncertainty of the diagnose eye disease with increase the accuracy of the system in determine the level of confidence of a disease that has been generated by the expert system. The expected results of this research is to provide a media consultation that is able to diagnose eye diseases and increasing the level of system confidence to diagnose an eye disease. This expert system is tested by comparing diagnose results from an expert and the comparasion results show accuracy rate is 96%.