When a decision needs to be made, other people’s perspectives might be significant, especially when time and money are at stake. People frequently rely on the prior experiences of their peers. Text analysis of people’s feelings, ideas, and emotions is used to analyze sentiments or opinions. It has recently emerged as the most active field of study for text mining and natural language processing. Thus, our objective is to demonstrate that using sentiment analysis techniques to do this is possible. The online product reviews on the Ali–Express e-commerce website provide the dataset used here. We employ a number of data processing methods, including stemming, TF-IDF, tokenization, parts of speech tagging, and the elimination of punctuation and stop words. Our study’s findings include a number of machine learning approaches. In this paper, we compared the precision and accuracy of eight different machine learning algorithm types, including the Naive Bayes Classifier, SVM, Random Forest Classifier, Logistic Regression (LR), K-nearest neighbors, XGBoost Classifier, Decision Tree Classifier, and Gradient Boosting. The results showed that these algorithms had the highest accuracy. Our investigation shows that, with an accuracy of almost 98%, logistic regression outperforms the other seven classifiers, while KNN performs the worst, with an accuracy of 46%. Other classifiers, like Random Forest and XGBoost, provide 80% accuracy, Decision trees provide 69%, Gradient Boosting 63%, Naive Bayes and SVM 82%, and so on.