Implementasi Algoritma C4.5 dan Naive Bayes untuk Analisis Sentimen Publik terhadap Platform Live Streaming Dukov di Indonesia

Abstract

The rise of online live streaming application platforms in Indonesia has led to various public opinions. As technology advances, live streaming applications not only provide live features, but also have many other features such as games, voice calls, video calls and many more. In this paper, it aims to analyze public sentiment towards the Dukov live streaming platform in Indonesia. In this study, the C4.5 and Naïve Bayes algorithms are used to analyze ratings and comments given by the public on the Google Play Store. In this study, the Naïve Bayes algorithm and the C4.5 algorithm are used to obtain results from user reviews into the appropriate sentiment category. By comparing the Naïve Bayes and C4.5 algorithms, this study aims to improve the accuracy and usefulness of sentiment analysis of Dukov live streaming application reviews. Based on 1000 review data divided into two categories, namely 300 training data and 700 testing data, then testing was carried out using 2 algorithms at once, namely C4.5 with the acquisition of 549 positive sentiments, 148 negative sentiments, and 3 neutral sentiments while the Naïve Bayes Classifier algorithm with the acquisition of 407 negative sentiments, 293 positive sentiments, and 0 neutral sentiments. The results of this sentiment analysis provide deeper insight into user views and assessments of the Dukov live streaming application.