Hybrid Recommender System Using A-priori Algorithm

Abstract

Recommendations have become an integral part of almost all information based as well as e-commerce systems. The aim of the recommendation systems is to harness the large information and product catalogues and understand the user’s preferences based on their choices, and recommend them products that otherwise would have been impossible to select through the massive product spaces manually. Research in this field has been able to identify a variety of algorithms and methodologies to make user-centric recommendations. Each algorithm covers a different set of parameters under consideration based on the requirement of specific tasks or a domain of products from the point of view of each user’s personalization and preferences. This research paper discusses the existing approaches used by various recommendation systems, their comparison, and proposes a method that addresses the shortcomings of existing practices in building recommendation systems by using the Apriori algorithm and employing association rules.

Keywords: Recommendation System, Content Based Filtering, Collaborative Filtering, User Preferences, Information Catalogues, User Ratings, Predictions, A-priori Algorithm, Association Rules