Preprocessing matrix factorization for solving data sparsity on memory-based collaborative filtering

Abstract

Collaborative filtering (CF) is one of the techniques in recommender system which utilizes information of user preference in the form of ratings of items and produce recommendation based on the similarity of behaviors with other user's preference. The collaborative filtering approach divisible into two main categories: memory-based and model-based, both have their respective advantages and disadvantages. The weakness of memory-based CF is that accuracy becomes less optimal when using sparse dataset. We propose the use of matrix factorization as preprocessing to fill empty rating values to handle sparse rating data. The research involves memory-based CF, with and without preprocessing to analyze both prediction performances. Our results show that the proposed approach with preprocessing has better accuracy than without preprocessing.