Expert System for Early Detection of Depression Using Psychological Symptoms Certainty Factor Method

Depressive disorders in the elderly often go undetected due to early symptoms that resemble normal aging processes. The absence of an early detection system becomes a major obstacle to prompt treatment. This study aims to design an expert system for early detection of depression in the elderly using the Certainty Factor (CF) method. The dataset was collected from 60 patient complaint narratives and validated by three professional psychologists with over five years of experience in geriatric psychiatry. The system design process includes symptom extraction using Natural Language Processing (NLP), CF value calculation for each symptom, and classification of depression risk (low, moderate, high). The system architecture consists of a knowledge base, inference engine, and user interface. Validation was conducted through diagnostic accuracy testing and user evaluation using a Focus Group Discussion (FGD). The results showed a validity level of 73%, and 88.6% of respondents agreed that the system can assist in early diagnosis. The novelty of this study lies in the integration of NLP and Certainty Factor tailored to the narrative patterns of the elderly, combined with a user-friendly interface design. This system is expected to serve as a supportive tool for psychologists and families in the early detection of depression in elderly individuals.