IMPLEMENTATION OF A VIRTUAL ASSISTANT AS AN INTERACTION FRIEND BY IMPLEMENTING NATURAL LANGUAGE PROCESSING: AUTOMATIC SPEECH RECOGNITION

Kurniawan, Dimas Rizqi (2026) IMPLEMENTATION OF A VIRTUAL ASSISTANT AS AN INTERACTION FRIEND BY IMPLEMENTING NATURAL LANGUAGE PROCESSING: AUTOMATIC SPEECH RECOGNITION. Tugas Akhir thesis, University of Technology Yogyakarta.

[img] Text
5210411399_Dimas Rizqi Kurniawan_Abstrak.pdf
Download (101kB)

Abstract

ABSTRACT Mental health problems remain a global concern, with Indonesia reporting relatively high rates of depression and anxiety. The lack of outlets for emotional expression often exacerbates these conditions, ultimately leading to a decline in mental well-being. Although artificial intelligence (AI)-based virtual assistants are commonly used in e-commerce and the education sector, their application in mental health support remains underexplored. Most current implementations rely on text-based, transactional interactions, limiting their ability to foster empathetic and natural communication. This study develops a voice-based virtual assistant that integrates Automatic Speech Recognition (ASR) for voice input, a generative AI model to produce empathetic responses via a large language model, and a Text-to-Speech (TTS) module trained on a specialised Indonesian dataset to adjust accent and prosody. The system is evaluated based on the relevance of responses and the naturalness of speech to ensure humanistic communication. This application aims to provide users with a safe space to express themselves without fear of stigma, support emotional well-being, and encourage early reflection on mental health conditions. Keywords: depression level, generative artificial intelligence, voice recognition, Tacotron2 Waveglow, virtual assistant

Item Type: Thesis (Skripsi, Tugas Akhir or Kerja Praktek) (Tugas Akhir)
Subjects: T Technology > T Technology (General) > T201 Patents. Trademarks
Divisions: Fakultas Sains Dan Teknologi > S1 Informatika
Depositing User: Kaprodi S1 Informatika UTY
Date Deposited: 05 May 2026 03:29
Last Modified: 05 May 2026 03:29
URI: http://eprints.uty.ac.id/id/eprint/19775

Actions (login required)

View Item View Item
UTY Open Access is powered by EPrints 3 which is developed by the School of Electronics and Computer Science at the University of Southampton. More information and software credits.