Perceptual Based Sentiment Analysis of Consumer Reviews using Rational Machine Learning Techniques for E-Commerce Applications

The growth of e-commerce has led to a surge in online shopping causing people to share their reviews and opinions on the product purchased. To help companies identify negative comments, the study uses natural language processing (NLP), K-means clustering, and sentiment analysis on a dataset of product reviews. The goal was to generate graphical and extractive summaries that highlight aspects that contribute to product failure. The system computed the sentiment polarity of each aspect and produced opinionated summaries. The system was evaluated on multiple review datasets and outperformed a baseline approach using Support Vector Machine (SVM) in terms of precision and recall.