A Hybrid Recommendation Model Based on Weighted Bipartite Graph and Collaborative Filtering

Recommender systems are designed to solve the overload information. Collaborative filtering based on the entire user network is by far most widely recommended algorithm, but it produced large amounts of operational data and it has difficulty to analyze characteristics of products and deal with data sparsity problem. To solve this problem, we employed a hybrid recommendation model which combined the weighted bipartite network with item based collaborative filtering. The experiment was implemented on dataset BookCrossing. Compared with traditional recommendation algorithms, the results proved that the proposed algorithm shows better performance and higher accuracy.