A Comparative Study on the Different Approaches Used for a Recommendation System for Games

Abstract: Recommender systems can be found in almost every domain in the current world. It is a multidisciplinary field, utilizing data mining and machine learning and some other similar techniques as per the domain. Be it a shopping site, media streaming platforms, while navigating with Google maps or even booking an appointment. In the current world of overloaded technology, users are bombarded with recommendations where ever one goes. Here the focus is on a game recommending system which suggests its users what game to buy next. The different approaches used for recommending games for a particular user is compared and contrasted. We see how the approaches have their own perks and losses. We take a look at the content-based filtering approaches for a game recommendation system and a collaborative filtering system. Also gives a closer look at a deep learning system to see if that bridges the gap between the content-based and collaborative approaches.

Keywords: Recommender systems, content-based filtering, collaborative filtering, hybrid methods,Deep learning, reinforcement learning.