A Collaborative Filtering Recommendation Algorithm Based on Item Similarity of User Preference
Tieli Sun, Lijun Wang, Qinghe Guo · 2009
The increasing users and items restrict the development of collaborative filtering recommendation systems. Then a series of problems, such as sparsity, cold start and scalability, come out. In this paper, we add user preference based on item genre, compute the similarity aimed at user preference. It can reduce the amount of data and improve the rapidity when computing similarity between items, and it can be more veracious and better recommendation quality. The experiment result shows that problems above can be solved with this approach.