Perbandingan Implementasi Metode Smote Pada Algoritma Support Vector Machine (Svm) Dalam Analisis Sentimen Opini Masyarakat Tentang Mixue

Abstract

In December 2022, the development of a franchise for an ice cream and tea outlet from China named Mixue became the talk of the Indonesian people, especially on social media Twitter, giving rise to various opinions from the public regarding the Mixue outlet which is growing so rapidly. So from that, sentiment analysis will be carried out by classifying using the implementation of the Support Vector Machine (SVM) algorithm. From the results of research that has been done, the SMOTE-based Support Vector Machine (SVM) algorithm results in an increase in the accuracy value to 73.67% and precision to 75.40%, and for the results of Support Vector Machine (SVM) without using SMOTE, the accuracy value is 69.40% and the precision value is 68.12%. But on the contrary there was a decrease in the recall value to 70.83% and the F1-Score value to 72.79%. So from the evaluation results it can be concluded that SMOTE has an effect on increasing the accuracy and precision values, but there is a decrease in the recall value and F1-Score.