IMPLEMENTASI FRAMEWORK REKAYASA KEBUTUHAN BERBASIS HUMAN-CENTERED AI PADA APLIKASI PENDAMPING EMOSI UNTUK KESEHATAN MENTAL

Riri Annisatunnaza, - and Mochamad Iqbal Ardimansyah, - and Asyifa Imanda Septiana, - (2025) IMPLEMENTASI FRAMEWORK REKAYASA KEBUTUHAN BERBASIS HUMAN-CENTERED AI PADA APLIKASI PENDAMPING EMOSI UNTUK KESEHATAN MENTAL. S1 thesis, Universitas Pendidikan Indonesia.

Abstract

Kesehatan mental remaja dan mahasiswa semakin mendapat perhatian seiring meningkatnya kasus gangguan mental emosional. Namun, stigma sosial, keterbatasan akses, dan kekhawatiran privasi sering membuat mereka enggan mencari bantuan profesional. Perkembangan kecerdasan buatan (AI) kemudian menghadirkan peluang baru melalui asisten virtual yang menyediakan ruang curhat mudah diakses. Penelitian ini menerapkan framework Requirements Engineering for Human-Centered AI (RE4HCAI) pada Kaola, sebuah aplikasi kesehatan mental berupa asisten virtual pendamping emosi berbasis Large Language Model (LLM) untuk remaja dan mahasiswa. Tujuan penelitian adalah mengidentifikasi kebutuhan pengguna, data, dan model dengan memperhatikan prinsip etika seperti privasi, transparansi, kontrol pengguna, keadilan, dan inklusivitas.Metode yang digunakan adalah Design Research Methodology (DRM), meliputi identifikasi kebutuhan melalui survei, pemodelan dengan pendekatan Goal-Oriented Requirements Engineering (GORE), pengembangan prototipe, serta evaluasi menggunakan User Acceptance Testing (UAT) dengan skala Likert dan WHO-5 Well-Being Index. Hasil penelitian menunjukkan bahwa Kaola memperoleh nilai rata-rata UAT sebesar 4,77 dari skala 5, menandakan tingkat penerimaan pengguna sangat tinggi. Sementara itu, pengukuran WHO-5 mencapai 79,47/100, yang menunjukkan kondisi kesejahteraan emosional responden berada pada kategori baik. Evaluasi berbasis prinsip HCAI menegaskan bahwa sistem tidak hanya memenuhi aspek fungsional tetapi juga sesuai dengan nilai etis dan kebutuhan pengguna. Dengan demikian, framework RE4HCAI dapat diimplementasikan secara fleksibe dalam mendukung pengembangan aplikasi kesehatan mental berbasis LLM yang relevan, adaptif, dan etis. ----------- The mental health of adolescents and college students is receiving increasing attention as cases of emotional and mental disorders increase. However, social stigma, limited access, and privacy concerns often prevent them from seeking professional help. Advances in artificial intelligence (AI) have presented new opportunities through virtual assistants that provide an easily accessible space for people to share their concerns. This study applies the Requirements Engineering for Human-Centered AI (RE4HCAI) framework to Kaola, a mental health application that functions as an emotional support virtual assistant based on a Large Language Model (LLM) for teenagers and university students. The research aims to identify user, data, and model requirements while considering ethical principles such as privacy, transparency, user control, fairness, and inclusivity. The method used is the Design Research Methodology (DRM), which includes requirement elicitation through surveys, modeling with the Goal-Oriented Requirements Engineering (GORE) approach, prototype development, and evaluation using User Acceptance Testing (UAT) with a Likert scale and the WHO-5 Well-Being Index. The results show that Kaola achieved an average UAT score of 4.77 out of 5, indicating very high user acceptance. Meanwhile, the WHO-5 Well-Being Index reached 79.47/100, showing that respondents’ emotional well-being falls into the good category. Evaluation based on HCAI principles also demonstrates that the system not only fulfills functional aspects but also aligns with ethical values and user needs. Thus, the RE4HCAI framework proves effective in supporting the development of LLM-based mental health applications that are relevant, adaptive, and ethically responsible.

	Text 58201_2100827_Title.pdf Download (526kB)
	Text 58201_2100827_Chapter1.pdf Download (279kB)
	Text 58201_2100827_Chapter2.pdf Restricted to Staf Perpustakaan Download (340kB)
	Text 58201_2100827_Chapter3.pdf Download (456kB)
	Text 58201_2100827_Chapter4.pdf Restricted to Staf Perpustakaan Download (522kB)
	Text 58201_2100827_Chapter5.pdf Download (269kB)
	Text 58201_2100827_Appendix.pdf Restricted to Staf Perpustakaan Download (1MB)

Item Type:	Thesis (S1)
Additional Information:	https://scholar.google.com/citations?view_op=new_articles&hl=en&imq=RIRI+ANNISATUNNAZA# ID SINTA Dosen Pembimbing : Mochamad Iqbal Ardimansyah : 6658552 Asyifa Imanda Septiana : 6681802
Uncontrolled Keywords:	Rekayasa Kebutuhan, Human-Centered AI, Kesehatan Mental, Nilai Etika AI, Asisten Virtual, Requirements Engineering, Human-Centered AI, Mental Health, AI Ethics, Virtual Assistant.
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science Q Science > QA Mathematics > QA76 Computer software
Divisions:	UPI Kampus cibiru > S1 Rekayasa Perangkaat Lunak
Depositing User:	Riri Annisatunnaza
Date Deposited:	30 Sep 2025 08:42
Last Modified:	30 Sep 2025 08:42
URI:	http://repository.upi.edu/id/eprint/136980

Actions (login required)

View Item