Product Recommendation System for E-Commerce using Collaborative Filtering and Textual Clustering

Abstract

In the past decade internet usage has seen an exponential growth. This growth has given opportunity for other businesses which rely on it to grow and prosper too. One such avenue is E-commerce. E-commerce is growing rapidly and with increase in popularity of E-Commerce, Recommendation has become very important too. Recommendation means providing relevant suggestions to the user as per his/her interest and need. Recommendations to users can be given based on price, area of living, wish listed items, cart items, searched items and previous purchased items. Recommendation systems improve user experience, boost sales and enhance user's engagement. Our proposed recommendation system will recommend products both for new users and existing users. This recommendation system uses model based collaborative filtering and recommends products based on rating and previous purchase history of the old users. Also the new users will get recommendations of new products, trending products and products on sale. Existing users will get recommendations based on recently viewed products, complementary products, etc. Since we're setting up a new e-commerce website, initially there are no user ratings for different products, so in this case, recommendations are made based on the textual clustering analysis of product description. Model based Collaborative Filtering along with Textual Clustering will help us in improving accuracy and targeting all types of users. E Commerce is gaining popularity and the recommendation system with E Commerce is like an icing on the cake.