CDSER: Sentiment Analysis for Product Selection to Enhance E-commerce Review System

Opinion mining is a widely booming genre where reasoning test using machine learning or lexicon-based. The basic goal of review generation is to keep the features of the text while shortening it without affecting the meaning of the content. The requirement for effective automatic review sparked significant interest in Opinion Mining and the Natural Language Processing areas. This research focuses on developing an extractive review method called CDSER is a candid sentiment analysis based on rule which generates grammatically combined words frequency. At first, we manage dictionary which contains words, phrase and associated polarity. Adjective detection method pick up most relevant words and combining method combines words grammatically and third method makes score on grammatically extracted words. We compare its effectiveness with two existing methods including VADER and TextBlob. The suggested technique has the advantages of increased computing efficiency, improved inferences from social media, data interpretation, resilience, and managing sparse data. Experiments on various datasets also surpass earlier research, and the accuracy is great, demonstrating the efficiency and innovation of the study report.