Skip to Main Content
Collaborative filtering has two methodologies: user based one and item based one. The former uses the similarity between users to predict, while the latter uses the similarity between items. Although both of them are successfully applied in wide regions, they suffer from a fundamental problem: data sparsity. In this paper, we propose a hybrid approach to overcome the problem. We define a similarity weight to dealing with the data sparsity. Experimental results showed that our new approach can significantly improve the prediction accuracy of collaborative filtering.