A Comparative Study of Amazon Product Reviews Using Sentiment Analysis

Abstract

Online shopping is an electronic business that allows people from all over the world to buy goods of their interest via web and various applications. Nowadays, these facilities are provided by famous E-commerce platforms such as Amazon, Flipkart, Snapdeal etc. Online shopping is one of the best businesses running over the Internet and hence it becomes the prime responsibility of these platforms to provide the best-rated products at the most feasible price. This paper provides a mechanism that can be used by various online shopping platforms to analyze the reviews given by the buyers, using sentimental analysis in order to maintain good service amongst their users. Sentimental Analysis is one of the most trending research areas in the domain of Natural Language Processing. It is defined as the technique that helps in the analysis of people’s emotions, sentiments from written text. In this paper, various Machine Learning classification algorithms have been used for finding the polarity of the reviews. Specifically, comparative analysis of algorithms such as Stochastic Gradient Descent, Logistic Regression, Multinomial Naive Bayes, and Support Vector Machine has been done. Performance evaluation of these algorithms has been done on the basis of the accuracy achieved. The observed results show that the Stochastic Gradient Descent with Bag of Words model outperforms other algorithms and shows the highest accuracy of 88.76%.