Smart Product Recommender System using Machine Learning

Abstract

To develop a comments based recommendation system using Recurrent neural network algorithm.  Recommendation systems have come a long way since collaborative filtering and content based systems. Considering comments as an important piece of date, which needs to be processed in order to extract information out of it and possibly combine its use cases with other recommendation systems. Since user comments are a direct form of data from the user it contains important keywords based on which further recommendations can be made. Here the source application is an ECommerce based app which proves to have a complex database in order to populate the data source. This algorithm can be implemented not only to an ECommerce application but also various different applications of the same kind where users and a service or product is involved. In today's world recommendations are the key to having a successful business. It is personalized and very effective that on a psychological level manipulates a customer to buy a particular set of products and also compares and recommends the best product rated by the users. The policy User recommendations by the User and for the User are followed.