Sentiment Analysis of Amazon Product Reviews: A Comprehensive Evaluation Using Naïve Bayes Classifiers

In the evolving landscape of digital commerce, sentiment analysis plays a pivotal role by computationally identifying and categorizing author sentiments in textual content. This practice has broad applications across industries, from predicting market trends to understanding customer satisfaction through online reviews and social media. The surge in online shopping has led to the dominance of e-commerce platforms like Amazon, where product prices may vary across multiple websites. Customers heavily rely on online reviews to gauge product worth, while manufacturers leverage these insights to meet customer needs and stay competitive. This study specifically explores the feasibility of applying sentiment analysis to classify product reviews from Amazon.com. Comparative analyses involve the application of different Naïve Bayes approaches. It has been observed that the Multinomial Naïve Bayes Model provides better accuracy as compared to other approaches, whereas the Complement Naïve Bayes Model is efficient in classifying both positive and negative sentiments.