Implementation of Collaborative Filtering in E-Commerce Recommender Systems: An Elementary Review

This paper presents an elementary review of product recommendation systems which is growing rapidly with significant market value. This paper explores different types of recommendation systems (RS) focusing mainly on collaborative filtering (CF) and other filtering techniques like content-based filtering (CBF), demographic-based (DB), utility-based (UB), knowledge-based (KB), and hybrid systems (HS). It summarizes the algorithms supporting these systems and their applications across cross domains focusing on the challenges as well. This paper outlines the qualitative metrics for evaluation and assessing the efficiency of a recommendation system's performance and provides a stepby-step guide for constructing a machine learning based recommendation system. The key challenges in collaborative filtering like the cold start problem, problem of data sparsity, and scalability issues are also addressed. Thus, the review aims to provide a detailed review of the current RS techniques, their implementation, and future growth in this field.