<![CDATA[Demonstrations against the Indonesian House of Representatives (DPR RI) have triggered extensive public opinion flows on social media; however, sentiment mapping of Indonesian-language comments on YouTube live broadcasts of political issues still requires more structured methodological reporting and evaluation. This study aims to classify public sentiment from 1,493 YouTube comments related to DPR RI demonstrations using the Support Vector Machine (SVM) algorithm. Data were collected via the YouTube Data API and subsequently processed through text cleaning, case folding, normalization, tokenization, stopword removal, and stemming. Sentiment labeling was performed using an Indonesian lexicon-based approach to generate three sentiment classes (positive, negative, and neutral), with neutral sentiment being dominant. Feature representation was constructed using CountVectorizer, and the SVM model was trained using an 80:20 split for training and testing data. Model performance was evaluated using accuracy, precision, recall, and F1-score metrics, achieving an accuracy of 92.4% (weighted performance of 0.924). Word frequency analysis was also employed to identify dominant terms within each sentiment class. These findings demonstrate the effectiveness of SVM in mapping digital public perceptions on political issues and highlight its potential to support data-driven policy evaluation.]]>
Copyrights © 2025
