一种基于用户行为相似度的协同推荐算法

he most crucial component of the algorithm is the mechanism of calculating similarities among users. Most existing calculations of similarities suffer from data sparsity and poor prediction quality problems. In this paper, a new CF algorithm based on User Behavior Similarity (UBS) is proposed, which extracts users' network behaviors' characteristics from the sparse rating matrix based on fully consideration of the features of data sparsity. By introducing a more realistic similarity measure, which is named user behavior similarity, the algorithm effectively solves the problem of the inaccuracy of similarities in data sparsity or small size neighborhood environments. Experiments show that algorithm outperforms other state-of-the-art CF algorithms and it is more robust against data sparsity.