A Switching Hybrid Approach to Improve Sparse Data Problem of Collaborative Filtering Recommender System
International Journal for Research in Applied Science and Engineering Technology
Collaborative filtering is a powerful technique that has been used in many recommender systems with considerable success. This technique uses the interested databases of users with items to predict what products they might like. Nevertheless, these interested databases collected very few under 10%, greatly affect the efficiency of the recommender system. Many research projects have given different solutions to solve this problem. Initially, they have made significant efficiency, improved
... cy, improved somewhat sparse data problem, but they still have existed their shortcomings. In this paper, we research and propose a new method for improving sparse data problem, which offers higher efficiency of recommendation than the approaching traditional collaborative filtering and some other methods. Keywords: Collaborative filtering, recommender system, sparse data, improve sparse data problem, hybrid recommender system.