The memory-based collaborative filtering algorithm is one of the most successful technologies for recom- mender systems, although these approaches all suffer from data sparsity and poor scalability problems. Folksonomy has emerged as a powerful mechanism that enables users to find, organize, and understand online entities. This paper focuses on how to address these problems by using tags. The experiment for the tag-based algorithm and other algorithms showed that the novel algorithm achieves better performance compared to the traditional ones, proving the validity of the algorithm.