Deteksi dan Analisis Sentimen Berdasarkan Aspek Gangguan Mental Menggunakan Bidirectional Gated Recurrent Unit dan Semantic Similarity

Sutranggono, Abi Nizar (2024) Deteksi dan Analisis Sentimen Berdasarkan Aspek Gangguan Mental Menggunakan Bidirectional Gated Recurrent Unit dan Semantic Similarity. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6025221051-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 April 2026. Download (3MB) | Request a copy

Abstract

Gangguan mental secara signifikan memengaruhi kehidupan sehari-hari dan termasuk salah satu penyebab utama bunuh diri. Meskipun telah banyak penelitian tentang deteksi gangguan mental di media sosial, fokus utamanya adalah mengidentifikasi keberadaan atau ketiadaan indikasi dalam postingan, dengan sebagian besar penelitian hanya berfokus pada satu gangguan mental tertentu, terutama depresi. Terdapat kurangnya analisis komprehensif terhadap hasil deteksi. Oleh karena itu, penelitian ini menganalisis gangguan mental secara lebih rinci dengan menerapkan deteksi dan analisis sentimen berdasarkan lima aspek, yaitu ADHD (attention-deficit hyperactivity disorder), anxiety, bipolar, depression, dan PTSD (post-traumatic stress disorder). Proses deteksi menggunakan bidirectional encoder representations from transformers (BERT) embedding dan model bidirectional gated recurrent unit (BiGRU). Selanjutnya, aspek kategorisasi menggunakan semantic similarity, yang menilai kemiripan antara istilah-istilah yang dihasilkan dari ekstraksi hidden topic melalui non-negative matrix factorization (NMF) dan kata kunci yang terkait dengan lima aspek gangguan mental, diekstraksi menggunakan kombinasi metode ekstraksi term. Selain itu, klasifikasi sentimen menggunakan BERT embedding dan model BiGRU. Metode yang diusulkan dapat mengidentifikasi gangguan mental, mengategorikan aspek, dan mengklasifikasikan sentimen dengan baik. Kinerja optimal tercapai dalam deteksi gangguan mental menggunakan BERT embedding dan BiGRU, dengan akurasi mencapai 0,9009. Aspek kategorisasi menggunakan semantic similarity dan BiGRU mencapai akurasi 0,8507, sementara klasifikasi sentimen melalui BERT embedding dan BiGRU mencapai akurasi 0,8717. Hasil analisis menunjukkan bahwa 90% teks yang terkait dengan gangguan mental menyampaikan sentimen negatif, dengan aspek depression mempunyai persentase sentimen negatif tertinggi, mencapai 25%.
=================================================================================================================================
Mental disorders significantly impact daily life and are among the leading causes of suicide. Despite numerous studies on detecting mental disorders on social media, the focus has primarily been on identifying the presence or absence of indications in posts, with most studies concentrating solely on one specific mental disorder, particularly depression. There is a lack of comprehensive analysis of the detection results. Therefore, this study analyzes mental disorders in more detail by applying detection and sentiment analysis based on five aspects, namely ADHD (attention-deficit hyperactivity disorder), anxiety, bipolar, depression, and PTSD (post-traumatic stress disorder). The detection process utilizes bidirectional encoder representations from transformers (BERT) embedding and the bidirectional gated recurrent unit (BiGRU) model. Subsequently, aspect categorization employs semantic similarity, which assesses the resemblance between terms generated from hidden topic extraction via non-negative matrix factorization (NMF) and keywords linked to the five mental disorder aspects, extracted using a combination of term extraction methods. Additionally, sentiment classification leverages BERT embedding and the BiGRU model. The proposed method can effectively identify mental disorders, categorize aspects, and classify sentiments. Optimal performance is achieved in detecting mental disorders using BERT embedding and BiGRU, with an accuracy of 0.9009. Aspect categorization, utilizing semantic similarity and BiGRU, attains an accuracy of 0.8507, while sentiment classification through BERT embedding and BiGRU reaches an accuracy of 0.8717. The analysis results indicate that 90% of texts related to mental disorders convey negative sentiments, with the aspect of depression having the highest percentage of negative sentiments, reaching 25%.

Item Type:	Thesis (Masters)
Additional Information:	RSDP 005.1 ABI d 2024
Uncontrolled Keywords:	analisis sentimen berdasarkan aspek, deteksi gangguan mental, semantic similarity, aspect-based sentiment analysis, BERT, BiGRU, mental disorder detection
Subjects:	B Philosophy. Psychology. Religion > BF Psychology Q Science > QA Mathematics > QA336 Artificial Intelligence R Medicine > R Medicine (General) > R858 Deep Learning T Technology > T Technology (General) > T57.5 Data Processing
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Informatics Engineering > 55101-(S2) Master Thesis
Depositing User:	Abi Nizar Sutranggono
Date Deposited:	01 Feb 2024 04:59
Last Modified:	21 Nov 2024 08:33
URI:	http://repository.its.ac.id/id/eprint/105874

Actions (login required)

View Item