Analisis Sentimen Terhadap Kinerja Wakil Presiden Pada Tahun 2025 Menggunakan Metode Support Vector Machine

Abstract

This study examines public perception of the performance of Indonesia’s Vice President in 2025 by utilizing opinion data from social media X/Twitter. The research addresses the lack of up-to-date quantitative insights into public sentiment polarity following the inauguration, particularly regarding Gibran Rakabuming Raka, whose appointment has sparked mixed reactions. The objective of this study is to classify sentiments as positive or negative and to evaluate the performance of the classification model on a corpus of user posts. The dataset consists of 898 tweets collected using the hashtags #wapres, #Gibran, and #WapresGibran. Data processing involved cleaning the text, converting all characters to lowercase (case folding), tokenization, normalization, removal of stopwords, and stemming. Feature representation was carried out using Term Frequency–Inverse Document Frequency (TF-IDF), while modeling was performed with the Support Vector Machine (SVM) algorithm. Results show 647 tweets with positive sentiment and 251 tweets with negative sentiment, indicating a generally positive tendency while maintaining some diversity of opinion. The SVM model achieved an accuracy of 80.68%, demonstrating reliable performance on high-dimensional textual data. These findings provide a concise overview of public opinion that can serve as a reference for policymakers and government communication strategies. The study’s main contribution lies in offering empirical evidence from social media on sentiment dynamics toward the Vice President’s performance, while also highlighting the effectiveness of combining TF-IDF and SVM in contemporary political sentiment analysis.