Analisis Sentimen Terhadap Istana Garuda Di Ibukota Nusantara (IKN) Menggunakan Algoritma Random Forest Dan Support Vektor Machine

Abstract

ABSTRACT

The relocation of Indonesia's capital city (IKN) to East Kalimantan is a national strategic project that has sparked diverse public opinions, particularly regarding the construction of Garuda Palace. This study aims to analyze public sentiment toward the Garuda Palace project using Random Forest and Support Vector Machine (SVM) algorithms and to compare their performance based on accuracy, precision, recall, and F1-score. This research offers three key novelties. First, it focuses on public opinion regarding the Garuda Palace project at IKN, which is underexplored in both local and international literature. Second, the use of Inset and Senti labeling techniques introduces a novel approach to sentiment categorization. Third, the comprehensive evaluation of Random Forest and SVM performance provides new insights into their effectiveness in large-scale infrastructure sentiment analysis in Indonesia. The methodology consists of five stages: (1) Data collection through web scraping from Twitter (July-August 2024) using keywords related to "Garuda Palace" and "IKN"; (2) Data preprocessing, including tokenization, stopword removal, stemming, and TF-IDF transformation; (3) Data labeling using Inset and Senti approaches; (4) Model training with Random Forest and SVM algorithms; (5) Model evaluation using confusion matrices and performance metrics such as accuracy, precision, recall, and F1-score. Results indicate that Random Forest achieved 77% (Inset) and 89% (Senti) accuracy, excelling in detecting negative sentiment with an F1-score of 0.93 on the Senti dataset. SVM achieved 89% (Inset) and 91% (Senti) accuracy, performing better in detecting positive sentiment with a precision of 0.96 on the Senti dataset. This study provides valuable insights into public perceptions of national infrastructure projects, supports data-driven decision-making, and serves as a reference for future sentiment analysis systems