Applying Matrix Factorization In Collaborative Filtering Recommender Systems

Collaborative filtering plays a vital part in advancing the recommendation environment by using the matrix factorization (MF) decomposition technology which is demonstrated to be most successful recommendation strategies. Despite being a successful method used in recommendation systems, SVD-based methods suffer from the data sparsity problem, which leads to inaccurate prediction of ratings. This paper proposes an incorporation-based recommendation method, to address the issue of sparsity in SVD-based strategies. Initially, similar users and items are found. Then, data is generated according to co-rated values. Finally, the data is incorporated into the SVD framework. We implemented our method on MovieLens 100k dataset. Experiment results represent that our method of prediction is better than the existing system.