Collaborative Filtering Algorithm Based on the Similarity of User Ratings and Item Attributes

Abstract

Collaborative filtering recommendation algorithm is key technologies of personalized re commendation system, as the serious data sparsity of rated items, the traditional collaborative filtering algorithms only depending on users data cannot achieve satisfactory recommended quality, an improved collaborative filtering recommendation algorithm based on the similarity of user ratings and item attributes is proposed. The experimental results based on Movie Lens dataset show that the improved hybrid collaborative filtering recommendation algorithm obtains the better recommendation accuracy than traditional similarity calculation method.