Sentiment and Customer Loyalty Analysis of Shopee Using Machine Learning Algorithms

Abstract

The exponential growth of e-commerce platforms has transformed consumer shopping behavior globally, including in Indonesia. Shopee, as one of the dominant online marketplaces, continuously attracts millions of active users through competitive pricing strategies, promotional events, and digital convenience. However, understanding user satisfaction and loyalty remains a challenge in such dynamic environments. This research aims to analyze user sentiment and customer loyalty toward Shopee by integrating computational sentiment analysis techniques with behavioral survey assessment. A total of 3,000 Shopee user reviews were collected through web scraping, then processed using text mining methods and classified into positive and negative categories using two machine learning algorithms: Support Vector Machine (SVM) and Naïve Bayes Classifier (NBC). Additionally, a structured loyalty survey was distributed to 30 respondents to evaluate behavioral loyalty indicators such as repeat purchase, advocacy, and emotional attachment. The SVM algorithm demonstrated superior performance with an accuracy rate of 98%, surpassing the Naïve Bayes Classifier’s 85% accuracy. The loyalty survey indicated a strong positive correlation between sentiment polarity and customer retention, revealing that satisfied users exhibit consistent repurchase intentions and brand advocacy. These findings emphasize the significance of integrating computational analytics and behavioral measurement in e-commerce performance evaluation. The results also provide managerial insights for enhancing digital service quality, consumer engagement, and long-term competitiveness in Indonesia’s online retail market