Collaborative Filtering Recommendation Algorithm Based on Improved Similarity Computing

At presently the most widely used algorithm is collaborative filtering in the Personalized Recommendation Systems for E-Commerce. Aiming at the problem that the recommendation is not accurate due to the data sparsity, a collaborative filtering algorithm based on user characteristics and item attributes preference was proposed in this study. It obtained the nearest neighbor users and similar items by analyzing the user characteristics, item attributes and the data of user's historical scores, and then computing the similarity between the two users based on the user-based collaborative filtering algorithm. It gave an algorithm which has the lower value of MAE and may improve the accuracy of the recommendation services.