Generating Positive and Negative Sentiment Word Clouds from E-Commerce Product Reviews

Abstract

Most customers who prefer buying products online on E-Commerce websites tend to rely on the ratings given to a product by other customers or a summary of the already existing customer reviews. However, a plethora of meaningful data is stored in the review text which eludes representation through customer ratings or the summary of the reviews likewise. But it is inefficient to go through each and every review. Our model thus adopts two approaches to demonstrate and resolve the generated issue - General Approach where the data is sorted based on the ratings, and Specific Approach where the data is sorted based on the products. The subsequent result is the generation of two new corpora followed by the generation of two new Word Clouds consisting of positive and negative features respectively for each existing product. The purpose of these Word Clouds is to highlight the features of products that are mentioned in the reviews. Hence, such a model provides more accurate as well as an efficient analysis of the offered products.