Recommendation Scheme via Improved Iteratively Collaborative Filtering Algorithm with Neighborhood Scale Research

Collaborative filtering algorithm is a traditional algorithm for recommendation system, and user rating (user-based collaborative filtering algorithm) is the primitive method of collaborative filtering algorithm. In this paper, with respect to the sparse data, the strategy of both user-based and item-based collaborative filtering algorithm is proposed and a three-steps updating algorithm is used, along with an iterative method to form a consistent recommendation scheme. Finally, the neighborhood scale is studied.