Reseach Review Abstract

With the exponential explosion of information growth, one of the recommended system techniques, collaborative filtering, has become a powerful active technique to utilize data in the information overload scene. This review first introduces collaborative filtering and its key disadvantages, then describes and compares two kinds of collaborative filtering: Memory-based approach and Model-based approach. Then we take a look at the evaluation metrics and prevalent datasets of collaborative filtering. Finally, we discuss the future challenges of collaborative filtering.