Recommendation System Using Neural Collaborative Filtering and Deep Learning

Abstract

Recommender systems have transformed the nature of the online service experience due to their quick growth and widespread use. In today’s world, the recommendation system plays a very vital role. At every point of our life, we use a recommendation system from shopping on Amazon to watching a movie on Netflix. A recommender system bases its predictions, like many machine learning algorithms, on past user behavior. The goal is to specifically forecast user preference for a group of items based on prior usage. The two most well-liked methods for developing recommender systems are collaborative filtering and content-based filtering. Somehow, we were using the traditional methods, named content-based filtering (CB) and collaborative-based filtering (CF), which are lacking behind because of some issues or problems like a cold start and scalability. The approach of this paper is to overcome the problems of CF as well as CB. We built an advanced recommendation system that is built with neural collaborative filtering which uses implicit feedback and finds the accuracy with the help of hit ratio which will be more accurate and efficient than the traditional recommendation system.