User-Item Recommendation System (UIRS) Using Collaborative Filtering

Abstract

The orthodox merchandise conduct has been upgraded by the e-commerce industry. Presently, people with busy routine, prefer online shopping instead of visiting physically to buy the products. In the modern era, topical prediction concerning a large amount of information regarding online shopping products is generating due to which users face the difficulty in finding the relevant information of products and services matching their tastes and preferences. A recommendation system is a powerful tool used by the e-commerce industry in order to assist e-buyers to showcase their products in an effortless and rapid manner. This paper initially discussed various collaborative filtering techniques. A proposed recommendation system is designed by using memory-based collaborative filtering techniques that are user-based and item-based filtering recommendation techniques. The proposed work focuses on using three items namely user name, item name, and item rating. The proposed model uses user rating data for the filtering of products and users. The system calculates the similarities between users and items using the Pearson correlation similarity measures for giving recommendations.

Keywords: e-commerce, collaborative filtering, model-based collaborative filtering, memory-based collaborative filtering, recommendation system.