Analisis Sentimen Publik Terhadap Pemilu 2024 di Media Sosial X Menggunakan Metode Text Mining

Abstract

The 2024 General Election marked a critical phase in Indonesia’s democratic dynamics, where social media played a central role as a platform for political expression and public opinion. X (formerly Twitter) emerged as a dominant arena for users to voice perspectives, support, and criticism toward electoral processes and political actors. This study aims to analyze public sentiment toward the 2024 election using a text mining approach and to identify dominant themes and sentiment fluctuations during the campaign and post-election period. A total of 87,350 tweets were collected through crawling, followed by preprocessing stages including cleansing, tokenization, TF-IDF transformation, and classification using Naive Bayes and Support Vector Machine (SVM) algorithms. The results show that negative sentiment dominated (45%), followed by positive (35%) and neutral (20%). SVM outperformed Naive Bayes, achieving an accuracy of 88.5%. Thematic and temporal analyses revealed strong public reactions to key events such as candidate debates and vote count announcements. This study highlights the value of social media as a real-time indicator of political opinion in the digital era.