Expert System for Detection of Diseases in Layers Using Forward Chaining and Certainty Factor Methods

Abstract

Inaccuracies in the process of diagnosing a type of disease result in errors in handling so that it will pose a risk of death. Accurate diagnostic process results require a high level of confidence so that the results are truly convincing. Current technological developments are making more and more mindsets for the development of information technology in the field of computerization born. One of them is an expert system. This expert system is often used to analyze disease in laying hens. The deficiency in previous research is that there is no degree of confidence so what happens is that the diagnosis often only uses the value of the expert. The role of the system user is only to select the available symptoms without giving the weighted value of the selected symptoms. This study aims to build an expert system capable of detecting symptoms in laying hens by assigning a degree of confidence to each symptom. The system is built with a combination of forward chaining techniques with a certainty factor, the weight value is based on a combination of the weight of symptoms from users and experts to anticipate conditions that are not ideal. Several stages in the research include data collection, knowledge base modeling, implementation into applications and testing. The conclusion that can be drawn from the trial results is that the system can show a maximum validity value of up to 100% when compared to manual calculations.