An Optimized Collaborative Filtering Approach with Item Hierarchy-Interestingness

Collaborative filtering algorithm is one of the most successful recommender technologies and has been widely adopted in recommender systems. However, the traditional collaborative filtering always suffers from sparsity problem of dataset. Item resource has hierarchy itself, and people´s interests are centralized in several hierarchies. In addition, rating is multivariate with several factors: user´s interest and item´s quality etc. The proposed algorithm makes corresponding modification based on the traditional algorithm with the ideas above. Experimental results show that the algorithm can guarantee the accuracy of the system recommended by the case, effectively alleviate the data sparsity problem.