Recommendation System for E-Commerce Using Memory Model Based Collaborative Filtering

Abstract

In today's technical era, every startup or a company attempt to establish a better sort of communication between their products and the users, and for that purpose, they require a type of methods which can promote their product effectively, and here the recommender system serves this motive with positivity. we all know numerous attempts have been made for expanding the accuracy in recommendation system, but somehow recommendation scenarios are much more complex and most of the cases have limited rating system for their items. Here, we present a demonstrative approach that will show how memory and model-based collaborative filtering enhance the accuracy and efficiently in our proposed recommendation system. As we know recommendation systems are used in many various areas’ like music, movie, news, books, social media platform. It is a filtering system that tries to predict and show the items that a user would like to purchase. In this paper we are using Memory based (Item to Item) and Model based (Item to User) collaborative filtering to solve various problem like cold start, grey sheep, data sparsity using Utility Matrix method which will help to find user item relationship from previous purchase history and cluster the Item to item relationship using K-means Algorithm which will solve the problem of cold start, grey sheep and spared the data problem.