Perbandingan Algoritma NaÏve Bayes, SVM dan Random Forest pada Analisis Sentimen Pengguna X terhadap IKN

Abstract

Sentiment analysis was explored to understand social media users' opinions towards the Indonesian Capital City (IKN) through the X platform with machine learning and lexicon-based algorithms. This research uses three algorithms: Naïve Bayes, Support Vector Machine (SVM), and Random Forest. The aim of this research is to test and compare the performance of the three algorithms to determine the best in classifying sentiment data from the X platform. The data consists of 10,000 tweets collected using the crawling method with the Python Harvest Library and Node.js, using keywords related to IKN. Based on the algorithm performance test, it was concluded that SVM had the highest performance compared to Naïve Bayes and Random Forest, producing an accuracy of 87%, precision 87%, recall 87%, and f-1 score 87%. This research uses the CRISP-DM Data Mining framework to ensure a structured and systematic approach to the analysis process.