A FRAMEWORK FOR GENERATING RANKINGS TO E-COMMERCE PRODUCTS BASED ON REVIEWS USING NLP

There has been a tremendous change in the way people buy and sell products/goods/services in the past decade, the modern way of shopping is done through E-commerce (Online Shopping) Amazon, Alibaba, flipkart are the current giants in E-Commerce. There has been a huge inclination towards online shopping due to various reasons such as wide-range, availability, various choices, easy to compare, can be brought/sold from anywhere, Unlike offline stores the consumer cannot verify the product before buying, the alternative way for this is relying on the reviews and ratings of the products. Which might be generated by various customers or spammers, as the range of the users cannot be confined there not accurate rating or review that can truly help the consumers. In this, we aim to overcome the trouble of avoiding such inaccurate data or insufficient data in terms of decision making, we aim to provide the users or customers a reliable ranking that help them to analyse the product and evaluate the product based on the ranking generated, We intend to create an algorithm that takes the reviews and generates an overall rank that should be helpful in decision making of the consumer. We plan to achieve this by using NLP (Natural Language Processing) techniques such as sentimental analysis and various NLP kits such as CNTK, KERAS Wrapper etc. for text mining for mining the review texts collected