The Design of Expert System for Determining the Initial Diagnosis of Tropical Infectious Diseases in Indonesia With Naive Bayes Method-Based Android

Abstract

Tropical infectious diseases are frequent, serious and concerning for the people in Indonesia. Tropical infectious diseases can be fatal and cause death. But if we diagnose them earlier and get proper treatment, the story can be changed. In this research will make a mobile application using Naive Bayes and Forward Chaining methods for early diagnosing tropical infectious diseases including typhoid fever, dengue fever, tuberculosis, malaria, and measles. The process of this application will start with input of the symptoms felt by users, after the data collected, system will calculate the data with Naive Bayes formula. This application using 147 data training from interviewed with the experts. Based on the tests by System Usability Scale method shows above average users rating 73.875 %, which means the results of the application are acceptable. And Confusion Matrix method shows performance of the application as high as 76.74 %.

Expert System for the Timely Diagnosis of Infectious Diseases

“…Regarding the method, the proposed expert system uses a rule-based methodology in terms of its operation. Instead, the systems proposed by the authors [11], [19] are based on the Naive Bayes method, and from [13], [14] is a certainty factor and in [20] is Bayes theorem.…”

“…Therefore the study of some solutions applied to the expert system for diagnosing infectious diseases is developed below. Permana et al [11] developed a mobile application for diagnosing infectious diseases such as typhoid, tuberculosis, dengue, measles, and malaria using Naive Bayes and Forward Chaining. After inputting the symptoms, the system infers the diagnosis using the Naive Bayes formula.…”