Syawalian Rais Dwi Jayanto, Suprihadi Suprihadi

Sentiment Analysis on X, TikTok, and Instagram on Indonesian Capital relocation using Support Vector Machine

Introduction

Sentiment analysis on x, tiktok, and instagram on indonesian capital relocation using support vector machine. Analyze public sentiment on Indonesia's new capital (IKN) across X, TikTok, & Instagram using SVM. Reveals platform-specific perceptions for effective policy communication strategies.

Abstract

This study examines public sentiment toward Indonesia’s new capital city, Ibu Kota Nusantara (IKN), across three major social media platforms: X, TikTok, and Instagram. The research aims to identify how public perceptions differ across platforms and to understand their implications for policy communication. A total of approximately 6,000 user comments collected up to March 2025 were processed through standard text-mining procedures, including cleaning, tokenization, stop-word removal, and stemming. The text data were converted into numerical features using the Term Frequency–Inverse Document Frequency (TF-IDF) technique and classified using a linear Support Vector Machine (SVM) model. Model evaluation with a 20% hold-out test set yielded an accuracy of 90.23% and a macro F1-score of 0.8905. The analysis shows that overall sentiment toward IKN is predominantly positive, with Instagram and TikTok generating more supportive narratives, while X displays a higher concentration of critical or negative comments. These findings highlight significant platform-specific differences that can inform more effective public communication strategies regarding the IKN project.

Full Text

Comments

You need to be logged in to post a comment.