Analisis Sentimen Ulasan Aplikasi TikTok Shop Seller Center di Google Playstore Menggunakan Algoritma Naive Bayes

Abstract

In the rapidly developing digital era, users' views on mobile applications are a key factor in the success of an application. Understanding user sentiment can help application developers and management to improve service quality and user satisfaction. One of the social media that is experiencing a revolution is TikTok, a short video sharing platform that presents e-commerce innovations through the TikTok Shop Seller Center. Therefore, sentiment analysis was carried out to find out whether user reviews of the TikTok Shop Seller Center application tended to be positive or negative based on the Naïve Bayes algorithm. The research methodology involves data scrapping, data cleaning, preprocessing (case folding, stopword removing, tokenization, stemming), labeling, TF-IDF, data testing using confusion matrix and visualization using wordcloud. The results of research regarding sentiment analysis of reviews of the TikTok Shop Seller Center application on Google Playstore totaling 5000 data, it was concluded that user reviews were classified as negative with a percentage of 86.3% accuracy value, 83.7% precision value, 94.6% recall value and 88.7% % F1-Score value.