Study of Recommender Systems Techniques (original) (raw)

Abstract— Recommender systems provide a way to make the user’s search for required data from a huge reservoir of data easier. This also benefits the E-learning and E-commerce, which host large databases with a large number of products. This paper attempts to study the basics of the recommender systems and understand the transitions in the trends of approaches like the individual approaches of content-based, collaborative, knowledge-based, utility-based and demographic and their combinations given by hybrid approaches. It mainly focuses on two most successfully used techniques - Collaborative Filtering and Hybrid Systems, as well as the superiority of the latter over the former. The recent developments in hybridization in the field of Recommender Systems are also analysed in an attempt to track their progress.