<![CDATA[Shopee is one of the most popular e-commerce platforms in Indonesia, with millions of individuals using it to buy and sell online. The number of user reviews on the Shopee app could be an essential metric in terms of the amount of consumer satisfaction with the services provided. The research intends to analyse the sentiment of Shopee’s customer reviews using the Naïve Bayes and Support Vector Machine (SVM) algorithms. In this work, the dataset used is the customer review data from Shopee app retrieved from Kaggle platform with a total of 9,561 reviews. The research process includes data selection, text pre-processing (Case Folding, Cleaning, Tokenising, Stopword Removal, Stemming), sentiment labelling based on rating into positive and negative class, data balancing with Synthetic Minority Over-sampling Technique (SMOTE), feature transformation using Term Frequency- Inverse Document Frequency (TF-IDF) and data splitting into training and testing set with 80:20 ratio. Naïve Bayes and Support Vector machine approaches were used to carry out the classification. Evaluation parameters used were accuracy, precision, recall, F1-score and confusion matrix. The experimental results reveal that the Naive Bayes technique has an accuracy of 89.73%, a precision of 93.92%, a recall of 84.96% and an F1-score of 89.22%. However, SVM approach achieved 90.37% F1 score, 87.27% recall, 90.70% accuracy and 93.70% precision. The assessment result indicates that SVM technique is better than Naïve Bayes strategy in sentiment classification of customer reviews in Shopee. High accuracy and F1 score. The study result shows that the SVM method is better than the Naïve Bayes method in Shopee customer reviews sentiment categorisation.]]>
