The experiment results argue that the algorithm efficiently improves sparsity of rating data,and promises to make recommendations more accurately than conventional CF algorithms.