A Sentiment Analysis of Customer Reviews of Two Products Sold on Amazon
Penn collection
Degree type
Discipline
Subject
E-Commerce
Funder
Grant number
Copyright date
Distributor
Related resources
Author
Contributor
Abstract
Now more than ever, product reviews are used to help consumers make purchasing decisions. This form of online content serves as a potential source of data for researchers to look into for new ways to make conclusions about what marketing tactics or product qualities are valuable in today’s world of e-commerce. This case study of two nearly identical smart TVs investigates the correlation between review sentiment and consumer satisfaction in regards to a few factors: star ratings, product attributes, and price. Through sentiment analysis it was determined that sentiment towards a product is significantly less polarized than the star rating a consumer gives. An emotion map (word cloud) was used for an observational analysis into what attributes are important for the products in this case study. Price, pending further studies, was determined to have an effect on review sentiment for the two products through regression analysis.