Designing a Parallel Hybrid and Commercial Movie Recommendation System

Abstract

Recommendation systems have gained tremendous popularity over the past few years. As we all know recommendation systems have provided a boon in the field of online shopping and other browsing portals. Recommendation systems use machine learning techniques to predict what the user may like based on his history of interaction with a system full of items. At present there are many approaches known to implement a recommendation system. These approaches have several algorithms with various efficiencies. The efficiency and the accuracy of the recommendation system solely depend on the algorithm used. Hence, we’ve designed, implemented, and successfully deployed a system that uses an ensemble of item-, and content-based recommendation systems. In this paper we are going to explain how the results from two algorithms which are run on Hadoop are combined to get more accurate movie recommendations.