Deteksi Dini Gangguan Mental Menggunakan Chatbot Dengan Algoritma BERT

Umar, Maryani (2024) Deteksi Dini Gangguan Mental Menggunakan Chatbot Dengan Algoritma BERT. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5023201009-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 April 2027. Download (5MB) | Request a copy

Abstract

Kesehatan mental merupakan aspek kesejahteraan yang krusial, namun gangguan kesehatan mental terus meningkat dan berdampak signifikan pada individu dan masyarakat. Terutama, anak-anak dan remaja semakin rentan terhadap masalah ini, dengan bunuh diri menjadi penyebab kematian kedua di kalangan remaja. Masalah ini juga memiliki dampak ekonomi global yang besar, sementara akses ke perawatan kesehatan mental yang tepat terbatas. Penelitian ini bertujuan untuk mengatasi masalah tersebut dengan mengembangkan chatbot berbasis web yang menggunakan algoritma BERT (Bidirectional Encoder Representations from Transformers) untuk deteksi dini gangguan kesehatan mental. Chatbot ini bertujuan untuk memberikan saluran komunikasi yang nyaman dan tanpa stigmatisme bagi individu yang membutuhkan bantuan kesehatan mental. Melalui analisis teks yang canggih, chatbot dapat memahami pesan pengguna dan memberikan respons yang relevan. Berdasarkan pengujian, sistem chatbot algoritma BERT telah menunjukkan kemampuan interaksi yang cukup baik dengan memberikan respon yang sesuai dengan input pengguna. Data diambil dari 17 subjek dengan kriteria yang beragam melalui pengujian langsung pada localhost, di mana subjek berinteraksi dengan sistem menggunakan alur yang melibatkan pengiriman pesan dan pengisian kuisioner. Hasil deteksi dari sistem direkap dan disimpan, termasuk faktor pemicu dan initial support berdasarkan hasil deteksi disorders. Evaluasi menunjukkan bahwa meskipun respon chatbot sudah cukup sesuai, terdapat kendala saat input pengguna tidak sesuai dengan kata kunci yang ada di database. Hal ini menyebabkan sistem memberikan respon netral emosional yang seringkali tidak relevan dengan konteks pesan.
=====================================================================================================================================
Mental health is a crucial aspect of well-being, yet mental health disorders continue to rise, significantly impacting individuals and society. Children and adolescents are particularly vulnerable to these issues, with suicide being the second leading cause of death among teenagers. This problem also carries substantial global economic implications, while access to adequate mental health care remains limited. This research aims to address these challenges by developing a web-based chatbot using the BERT (Bidirectional Encoder Representations from Transformers) algorithm for early detection of mental health disorders. The chatbot seeks to provide a comfortable and stigma-free communication channel for individuals in need of mental health support. Through advanced text analysis, the chatbot can understand user messages and provide relevant responses. Based on testing, the BERT algorithm-based chatbot demonstrated satisfactory interaction capabilities by delivering appropriate responses to user inputs. Data was collected from 17 subjects with diverse criteria through direct testing on localhost, where subjects interacted with the system by sending messages and completing questionnaires. The system's detection results were summarized and stored, including triggering factors and initial support based on the detected disorders. Evaluation revealed that while the chatbot's responses were generally adequate, challenges arose when user inputs did not match keywords in the database. This limitation led to the system providing neutral emotional responses that were often irrelevant to the context of the messages.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Gangguan Mental, Stigma, Chatbot, Algoritma BERT Mental Disorders, Stigma, Chatbot, BERT Algorithm
Subjects:	R Medicine > R Medicine (General) > R857.M3 Biomedical materials. Biomedical materials--Testing. R Medicine > RZ Other systems of medicine T Technology > T Technology (General) > T58.6 Management information systems
Divisions:	Faculty of Electrical Technology > Biomedical Engineering > 11410-(S1) Undergraduate Thesis
Depositing User:	Maryani Umar
Date Deposited:	07 Feb 2025 03:20
Last Modified:	07 Feb 2025 03:20
URI:	http://repository.its.ac.id/id/eprint/118512

Actions (login required)

View Item