Analisis Perbandingan Algoritma Random Forest, Support Vector Machine, dan Naive Bayes dalam Klasifikasi Tingkat Pemahaman Radikalisme Pengguna Media Sosial

Abstract

The rapid growth of social media in Indonesia has significantly influenced communication patterns and information dissemination, including content that potentially contains radical ideologies. The high intensity of digital interactions exposes social media users to various ideologies quickly and extensively, creating an urgent need for objective and efficient methods to identify users’ levels of understanding of radicalism. This study aims to analyze and compare the performance of Random Forest, Support Vector Machine (SVM), and Naive Bayes algorithms in classifying levels of radicalism understanding based on questionnaire data. The research methodology includes questionnaire data collection, data preprocessing, feature extraction, model training, and performance evaluation using accuracy, precision, recall, and F1-score metrics. The results show that the SVM algorithm with a linear kernel achieves the best performance, with an accuracy of 95.83%, followed by Random Forest and Naive Bayes, each obtaining an accuracy of 72.92% and 70.83%. Confusion matrix analysis indicates that the SVM model minimizes misclassification across all classes, while feature importance analysis reveals that ideological beliefs and intolerant attitudes are the most dominant factors in determining levels of radicalism understanding. These findings demonstrate that machine learning approaches based on questionnaire data are effective for classifying radicalism understanding and have strong potential for development as early detection systems using data mining techniques.