Sentiment Analysis for Amazon Product Reviews Using Logistic Regression Model

Abstract : The sentiment is something that expresses the feeling about anything. This work proposes a Machine Learning approach that performs the classification of customer product reviews by finding the sentiment of the review dataset making it easy for product owners to track and improve their quality of the product rather than going through each piece of ratings and review. Amazon Product Reviews dataset is used for training and predicting systems, aiming to achieve maximum accuracy from the algorithm. This work also uses two different modules like Count Vectorizer and Tf-IDF Vectorizer for comparing the accuracy and consistency in training the text dataset, to conclude the better one between them.