Analisis Sentimen Era Kepelatihan Shin Tae Yong Dengan Menggunakan SVM dan SMOTE

Abstract

This study explores the application of a combined approach using TF-IDF, SMOTE, and Support Vector Machine (SVM) to address sentiment classification on imbalanced text data. The dataset consists of 3,377 social media reviews categorized into three sentiment classes positive, negative, and neutral. Text features were extracted using TF-IDF, and class imbalance was handled using the SMOTE technique. The SVM model was trained and evaluated, achieving an accuracy of 90.82% and a weighted average F1-score of 0.91. The results demonstrate that the proposed method effectively improves sentiment classification performance, particularly in handling class imbalance.