Traditional collaborative filtering is widely used in social media and e-business, but data sparsity and noise problems have not been solved effectively yet. In this chapter, we propose a novel approach of collaborative filtering based on field authorities, which achieves genre tendency of items by mapping tags to genres and simulates a fine-grained word-of-mouth recommendation mode. We select the nearest neighbors from sets of experienced users as field authorities in different genres and assign weights to genres according to genre tendency. Our method can solve sparsity and noise problems efficiently and has much higher prediction accuracy. Experiments on MovieLens datasets show that the accuracy of our approach is significantly higher than traditional user-based kNN CF approach in both MAE and precision tests.
展开▼
