Sentiment analysis on E-Marketplace User Opinions Using Lexicon-Based and Naïve Bayes Model

E-marketplaces experienced a surge in visits from 2018 to the present. The largest e-marketplaces in Indonesia (Tokopedia, Shopee, and Lazada) currently use operator services or e-services using Instagram and Twitter as customer service centers. Customer service centers have caused these three E-Marketplaces to receive massive user opinions on social media. Therefore, it is necessary to research a model that can understand the emotions in E-Marketplace users' opinions using sentiment analysis. This study uses a lexicon-based method to understand textual data in a dataset containing user opinions, then classify it using the Naive Bayes Model. This study examines, compares several models, and then evaluates them. This research uses the Lexicon-Based Model and Naive Bayes model. Thus, this research contributes to recommending the best model for Sentiment analysis on E-Marketplace User Opinion. The results of the study found that the combination of the TextBlob lexicon method and the Multinomial Naïve Bayes classification method outperformed the other three models.