Sentiment Analysis and Topic Modeling of E-Grocery Application Reviews Using Naive Bayes and Support Vector Machine: A Case Study of Segari Data Review on the Google Play Store
Jefka Dhammananda, Indra Budi, Aris Budi Santoso
Segari is a customer-centric company with a core value of being obsessed with its customers. The lack of human resources and the abundance of customer reviews that need to be analyzed hinder the process of extracting information from these reviews. Therefore, a machine learning model is needed to automatically perform sentiment analysis and topic modeling. The information extracted from sentiment analysis can be used as a reference to maintain service quality based on positive sentiments, while the results of negative sentiments can be used for evaluation to improve Segari's services and application. The data used on this research were customer reviews from the Google Play Store. The model development process includes data collection, data labeling, data preprocessing, feature extraction, sentiment classification model, model evaluation, and topic modeling. The researcher utilized two…