Astori, Muhammad Ichwan (2026) RANCANG BANGUN SISTEM KLASIFIKASI SENTIMEN KOMENTAR YOUTUBE MENGGUNAKAN METODE SUPPORT VECTOR MACHINE (SVM). S1 thesis, Universitas PGRI Madiun.
![[img]](https://eprint.unipma.ac.id/style/images/fileicons/text.png) |
Text
HALAMAN DEPAN.pdf Download (658kB) |
|
Text
ABSTRACT.pdf Download (10kB) |
|
Text
BAB 1.pdf Download (24kB) |
|
Text
BAB 2.pdf Download (4MB) |
|
Text
BAB 3.pdf Restricted to Repository staff only Download (97kB) |
|
Text
BAB 4.pdf Restricted to Repository staff only Download (5MB) |
|
Text
DAFTAR PUSTAKA.pdf Download (134kB) |
|
Text
BAB 5.pdf Download (12kB) |
|
Text
LAMPIRAN.pdf Restricted to Repository staff only Download (35kB) |
Abstract
This study aims to design and develop an aspect-based sentiment classification system for Indonesian public opinion regarding the Middle East conflict in YouTube comments. The limitations of traditional sentiment analysis in processing ambiguous compound sentences are addressed using the Aspect-Based Sentiment Analysis (ABSA) approach, covering humanitarian, economic, and political aspects. The system development utilizes the Agile method with the Django framework based on a Model-View-Template (MVT) architecture. Data was collected via external web scraping and uploaded in CSV format. Text transformation involved preprocessing (Sastrawi), Term Frequency-Inverse Document Frequency (TF-IDF) feature extraction, and handling class imbalance using the Synthetic Minority Over-sampling Technique (SMOTE). A linear kernel Support Vector Machine (SVM) algorithm was implemented as the classifier. Black-Box functional testing results showed that the system operated at 100% without errors. Based on Confusion Matrix testing on 1,628 opinion data, the SVM predictive performance achieved an Overall Accuracy of 88% for sentiment and 72% for aspect detection. In conclusion, the SVM algorithm is highly robust in dissecting high-dimensional compound opinions, and this system successfully mapped public perception granularly. However, this study has limitations, namely a prediction bias in detecting the minority class (positive sentiment), indicated by a Recall of 0.36 and an F1-Score of 0.48, as a result of the high data distribution imbalance in the testing corpus, as well as the need for additional slang lexicon integration.
| Item Type: | Thesis/Skripsi/Tugas Akhir (S1) |
|---|---|
| Kata Kunci: | Kata Kunci: Analisis Sentimen, ABSA, Support Vector Machine, Django, Konflik TimurTengah. Keywords: Sentiment Analysis, ABSA, Support Vector Machine, Django, Middle East Conflict. |
| Subjects: | T Technology > T Technology (General) |
| Divisions: | Fakultas Teknik > Teknik Informatika |
| Depositing User: | ASTORI ICHWAN MUHAMMAD |
| Date Deposited: | 17 Oct 2026 03:39 |
| Last Modified: | 17 Oct 2026 03:39 |
| URI: | http://eprint.unipma.ac.id/id/eprint/10235 |
Actions (login required)
 |
View Item |
