ANALISIS SENTIMEN KOMENTAR PADA TIKTOK SHOP MENGGUNAKAN METODE TF-IDF DAN KLASIFIKASI NAÏVE BAYES (STUDI KASUS: GLAD2GLOW)

Abstract

The rapid growth of social media and social commerce platforms has generated a large volume of user comments that can be utilized to understand consumer perceptions of a product. TikTok Shop, as one of the leading digital shopping platforms, contains various user comments expressing positive, negative, and neutral sentiments, making an automatic sentiment analysis approach necessary. This study aims to analyze user sentiments toward Glad2Glow products on TikTok Shop by applying the Term Frequency-Inverse Document Frequency (TF-IDF) method for feature weighting and the Naïve Bayes algorithm for sentiment classification. The dataset used in this study consists of 350 user comments collected from TikTok Shop. The research process includes data collection, text preprocessing, feature extraction using TF-IDF, sentiment classification with the Naïve Bayes algorithm, and model evaluation using a confusion matrix. The experimental results show that the classification model achieved an accuracy of 98.06%, with precision values of 1.00 for the negative class, 0.89 for the neutral class, and 1.00 for the positive class. The recall values were 0.92, 1.00, and 1.00, while the F1-scores were 0.96, 0.94, and 1.00, respectively. Furthermore, the web-based sentiment analysis system successfully performed automatic sentiment classification and displayed the classification results effectively. These findings indicate that the TF-IDF method and the Naïve Bayes algorithm provide excellent performance in classifying user sentiments toward Glad2Glow products on the TikTok Shop platform.