论文部分内容阅读
Nowadays,with the gigantic increment of users and items,traditional Collaborative Filtering (CF) suffers from the sparsity problem,which has become a key issue that affects recommendation quality.Aiming at alleviating this issue,we propose a novel CF algorithm named Genre-based Hybrid Collaborative Filtering (GHCF).Our algorithm improves the recommender accuracy in the following three crucial aspects:First,we introduce the genre information of items into traditional CF and transform the original sparse user-item matrix into a dense user-genre matrix.The user similarity calculation based on user-genre matrix overcomes the potential inaccuracy caused by the sparsity problem.Second,we address a new missing data prediction strategy by using item-based CF to fill vacancies when the rating of target users neighbor is missing during user-based CF process.We also build a user Recently Interested Genre Cloud (RIGC) for each user in order to tracking their interest more accurate.The neighbor users ratings are combined by weighed similarities and user recent interest to make the final prediction.Experiments on the MovieLens show that our GHCF algorithm can relieve the sparsity problem and effectively improve prediction accuracy compared with traditional CF.