Sentiment classification on product reviews using machine learning and deep learning techniques

Online product analysis is a frequently used tool that allows consumers to understand their needs readily. Every day, the selling and purchasing processes continue in an e-commerce store, and customer feedback keeps growing. Comments made by customers will serve as an evaluation of a product that customers have purchased. Customers can freely submit reviews with both positive and negative feedback in the e-commerce website’s Comments section. The authors will study the above concerns, utilising the opinion analysis technique to differentiate between the positive, negative, and natural product review categories and using machine learning and deep learning methods like LSTM, GRU, Support Vector Machine, BiLSTM, Random Forest, and CNN. Word clouds make comparing the three sentiment classifications in our research easier. Our findings demonstrate how sentiment analysis may be used to pinpoint customer behaviour, mitigate risk factors, and meet consumer expectations. The findings of our experiment show that the Random Forest method will produce superior outcomes than other currently used techniques.