A Hybrid Recommendation Method For Movies Using The Bisect K-Means And The Parallel SGD Algorithm

Abstract

A Recommender System is an approach for providing adequate information to the customer which speeds up the decision making process related to a particular task of interest. It is common, the taste of interest varies from person to person, and hence it is a challenge to design a recommender system which provides the suggestions based on individuals interest. Most of the current researches have given focus on either implementing the recommender system either by using content based or collaborative approaches in their research. As the effectiveness of the recommendation system depends on the accuracy and the speedy results, there is an ample space for the improvisation through upgrades in such systems.  In this paper, we have proposed a hybrid movie recommender system for the user by combining the better features of  Bisect K-Means and Parallel Stochastic Gradient Descent (SGD) algorithms. The experimental results have shown the considerable improvement over the approaches currently are in use