Machine Learning Based Recommendation System: A Review

Abstract

The digital era has created an extreme choice paradigm with an explosion of endless content. A user who is just starting on the platform or looking for a creature can be lost in this ocean. Therefore, it is necessary to design a system that can guide users as per their interest. To overcome this problem, the Recommendation System (RS) came into existence. RS is a tool used to recommend items as per user’s interests. The benefits of the RS cannot be exaggerated, given the potential impact to improve many of the problems associated with widespread use and over-selection in many web applications. In recent years, Machine learning (ML) shows great interest in different research areas, such as computer vision and Natural Language Processing (NLP), not only because of its stellar performance but also because of its attractive feature of demonstrating learning from scratch. The effect of ML techniques can be seen while applying these techniques to the prediction and recommender system. This paper presented a comprehensive survey on recommendation techniques used in conjunction with the ML approach in many domains. This work aims to find the shortcoming of available RS for different fields and the areas that require more effort to attain higher accuracy.