Mobile-Based Application Implementing the Dempster-Shafer Method for Diagnosing Post-Traumatic Stress Disorder (PTSD)

Abstract

Post-Traumatic Stress Disorder (PTSD) is a serious mental health condition that can develop after an individual experiences a traumatic event. Early diagnosis of PTSD symptoms is crucial to prevent more severe psychological and social impacts. This study aims to develop a mobile-based expert system application that implements the Dempster-Shafer method to independently diagnose the confidence level of PTSD symptoms. The system is designed with three main components: biodata input, a symptom selection interface based on 9 indicator questions, and an inference engine utilizing the Dempster-Shafer method with a Likert scale to assess the severity of trauma experienced by respondents. Testing was conducted on 47 respondents, and the analysis results showed that 19.1% were in the safe category, 4.3% in mild PTSD, 12.8% in moderate PTSD, and 63.8% in severe PTSD. Validity testing indicated that all symptom items had correlation values (r-calculated) greater than 0.2876, confirming their validity, and the instrument’s reliability was measured at α = 0.806. The application is capable of providing an initial diagnosis along with recommended actions based on the severity level of PTSD. This study demonstrates that applying the Dempster-Shafer method in a mobile application offers an effective, efficient, and accessible alternative solution for early PTSD diagnosis, especially for individuals with limited access to professional mental health services.