Sentiment Analysis Based Product Recommendation System

Abstract

E-commerce sites are now a day having boom in selling and purchasing products, some of the most popular e-commerce sites are Amazon, flipkart etc where most of the customers visit to purchase products. Product reviews (genuine comments from the customers) is very important for customers, seller, businesses and manufacturers. Seller often want to know in time what consumers and the public think of their products and services. However, it is not really very feasible to manually read every post on the website and extract useful viewpoint information from it because there are so many comments about a single product on any e-commerce site. If you do it manually, there is too much data and it consumes your huge time. Sentiment analysis allows large-scale processing of data in an efficient and cost-effective manner to analyze the sentiments and conclude the result. In order to explore more about sentiment analysis, this paper tries to use the power of sentiment analysis to help the buyer and seller both. It helps buyers to see the honest reviews of the customers which already had purchased some product in the past and also to the seller to analyse the bunch of reviews which is being posted by the customers who purchased the product and based on this analysis they can improve their product for better selling.