An Improved Collaborative Filtering Algorithm Based on User Clustering
Liwei Zhang · Information Sciences · 2014
Collaborative filtering algorithms have been successfully applied to the network personalizationrecommendation system, but data sparseness affects the recommendation quality of collaborative filteringalgorithms seriously. To solve this problem, this paper introduces the user activity and proposes animproved collaborative filtering algorithm based on user clustering. It extends user-item scoring matrixand improves user similarity computing method to alleviate the impact of data sparseness inrecommendation algorithm. So it can significantly improve the accuracy of the collaborativerecommendation algorithm. The experimental results show that it can describe the user similarityaccurately and improve the accuracy of recommendation algorithm.