Film recommendation systems using matrix factorization and collaborative filtering

Collaborative filtering method was widely used in the recommendation system. This method was able to provide recommendations to the user through the similarity values between users. However, the central issues in this method were new user issue and sparsity. This paper discusses about how to use matrix factorization and nearest-neighbour in film recommendation systems. Both of methods will be used in order to make more accurate recommendations. Based on the experiments results, the combination of matrix factorization and classical collaborative filtering (nearest neighbor) could improve the prediction accuracy. It can be concluded that the combination of matrix factorization and nearest-neighbor produced a better prediction accuracy.