Aplikasi Deteksi Dini Kesehatan Mental Menggunakan Metode Certainty Factor

Abstract

Abstrak

Kesehatan mental merupakan isu global yang memerlukan perhatian serius, terutama di kalangan mahasiswa dan masyarakat umum. Keterbatasan akses layanan profesional dan stigma sosial menjadi hambatan utama deteksi dini gangguan kejiwaan. Penelitian ini mengembangkan Aplikasi Mobile Deteksi Dini Gangguan Kejiwaan dengan metode Certainty Factor (CF) dan memvalidasi sistem menggunakan dataset mental health dari Kaggle. Model pengembangan menggunakan pendekatan Waterfall yang mencakup analisis kebutuhan, perancangan sistem, implementasi, pengembangan aplikasi menggunakan bahasa pemrograman Java dengan tools Basic4Android (B4A), serta pengujian dan evaluasi. Aplikasi dirancang dengan lima menu utama: Informasi, Mulai Tes, Riwayat, Konsultasi, dan Tentang. Dataset "Student Depression Dataset" dari Kaggle dengan 36,200+ records digunakan untuk validasi, mencakup variabel akademik, gaya hidup, dan faktor psikososial. Metode CF menghitung tingkat keyakinan diagnosis dengan mengombinasikan bobot pakar dan input pengguna melalui kuesioner dengan tiga pilihan jawaban (Tidak, Kadang, Sering). Sistem mampu mendiagnosis tujuh jenis gangguan kejiwaan: Skizofrenia, Gangguan Kecemasan, Depresi, Bipolar, OCD, BPD, dan PTSD. Hasil pengujian menunjukkan akurasi sistem 80.6%, dengan sensitivitas 68.4% dan spesifisitas 84.2%. Validasi dengan dataset Kaggle memperkuat reliabilitas sistem dalam mengidentifikasi faktor risiko. Penelitian ini berkontribusi pada pengembangan e-health untuk deteksi dini kesehatan mental yang accessible, user-friendly, dan evidence-based.

Kata Kunci: Kesehatan Mental, Certainty Factor, Sistem Pakar, Aplikasi Mobile, Basic4Android

Abstract

Mental health is a global issue requiring serious attention, particularly among students and the general population. Limited access to professional services and social stigma are major barriers to early detection of mental disorders. This study develops a Mobile Application for Early Detection of Mental Disorders using the Certainty Factor (CF) method and validates the system using mental health datasets from Kaggle. The development model employs a Waterfall approach encompassing requirements analysis, system design, implementation, application development using Java programming language with Basic4Android (B4A) tools, as well as testing and evaluation. The application is designed with five main menus: Information, Start Test, History, Consultation, and About. The "Student Depression Dataset" from Kaggle with 36,200+ records is used for validation, covering academic variables, lifestyle, and psychosocial factors. The CF method calculates diagnostic confidence levels by combining expert weights and user input through a questionnaire with three response options (No, Sometimes, Often). The system is capable of diagnosing seven types of mental disorders: Schizophrenia, Anxiety Disorder, Depression, Bipolar, OCD, BPD, and PTSD. Testing results demonstrate a system accuracy of 80.6%, with sensitivity of 68.4% and specificity of 84.2%. Validation with the Kaggle dataset strengthens the system's reliability in identifying risk factors. This research contributes to the development of accessible, user-friendly, and evidence-based e-health for early detection of mental health.

Keywords: Mental Health, Certainty Factor, Expert System, Mobile Application, Basic4Android