Implementasi Algoritma Naïve Bayes Terhadap Klasifikasi Ulasan Aplikasi Tokopedia

Abstract

Tokopedia has an average number of visitors per month which is 149.67 visitors. In addition, the Tokopedia application has been downloaded by more than 100 million users with 6.34 million reviews and received a rating of 4.7 out of 5. This study aims to classify the positive and negative reviews of the Tokopedia application on the Google Playstore. The results of the 2000 test data testing obtained 842 positive reviews and 1158 negative reviews. This means the percentage for positive reviews is only 42.1%, in contrast to negative reviews of 57.1%. The performance generated in this test against 2000 testing data is 96.19% accuracy value, with a precision value of 1. While in Class Recall the resulting value is 93.45% (positive class: negative). Then for the AUC value is 0.975.