author(s)
Kallal Banerjee and Sweety Sarkar

abstract
Nowadays, the world is becoming digitalized e-commerce is ascending in this digitalized world through the availability of products within reach of customers. Furthermore, e-commerce websites allow people to convey their thoughts and feelings. People are increasingly relying on the experiences of other customers. Our opinions and purchasing decision-making are affected by the experience of others and their feedback about products. We always ask others about their opinion to get benefit from their experience; hence, the importance of reviews has grown. However, it is almost impossible for customers to read all such reviews; therefore, sentiment analysis is essential in analyzing them. This study proposes a sentiment analysis to predict the polarity of Amazon baby product dataset reviews using supervised machine learning algorithms. Further, it will allow companies to improve their products by knowing customers’ opinions and needs. Amazon is one of the e-commerce giants that people use daily for online purchases where they can read thousands of reviews dropped by other customers about their desired products. These reviews provide valuable opinions about a product such as its property, quality, and recommendations, which helps the purchasers understand almost every detail. This project considers the sentiment classification problem for online reviews using supervised approaches to determine the overall semantics of customer reviews by classifying them into positive and negative sentiments.