A hybrid recommender system

Abstract

In the current era, the rapid pace of data volume is producing redundant information on the internet. Predicting the appropriate item for users has been a great challenge in information systems. As a result, recommender systems have emerged in this decade to resolve such problems. Many e-commerce platforms such as Amazon and Netflix are using some decent recommender systems to recommend their items to the users. Previously in the literature, multiple methods such as Matrix Factorization, Collaborative Filtering have been implemented for a long time, however in recent studies, neural networks have shown promising improvement in this area of research. In this research, motivated by the performance of hybrid systems, we propose a hybrid system for recommendation purposes. In the proposed system, the users are divided into two main categories: Average users and Non-average users. Both of these categories contain the users having similar behaviors towards the items. Various machine learning and deep learning methods are implemented in both of these categories to achieve better results. Machine learning algorithms such as Decision Trees, Support Vector Regression, and Random Forest are applied to the average users. For the non-average users, multiple techniques such as Matrix Factorization, Collaborative Filtering, and Deep Learning methods are implemented. The performed approach achieves better results than the traditional methods presented in the literature