Expert System for Diagnosing Behavioral Disorders Using the Dempster-Shafer Theory Algorithm

Abstract

Behavioral disorders, especially in children, can have a significant impact on their development in various domains, such as social, emotional, and academic. The diagnostic process for this disorder can be complicated due to overlapping symptoms and the possibility of complex causal factors. Early identification and appropriate treatment of behavioral disorders is essential to prevent more serious impacts on an individual's psychological well-being. Lack of knowledge and difficulty in accessing psychiatrists to find out about behavioral disorders in children are factors that result in this problem requiring a solution. The aim of this research is to create an expert system that utilizes the Dempster-Shafer Theory algorithm to detect behavioral disorders, thus simplifying the diagnosis process and ensuring accurate results. The Dempster-Shafer theory, as an inference engine, can overcome uncertainty by combining several sources of evidence or data that may overlap or be incomplete, resulting in a stronger conclusion. The main feature of this expert system is its ability to carry out diagnoses based on symptoms and display diagnosis results, disease descriptions, and treatment options. Test accuracy produces a value of 90%, which shows that the Dempster-Shafer Theory approach can diagnose behavioral disorders effectively.