抽象的な

User integrated similarity based collaborative filtering

Tian-Shi Liu, Nan-Jun Sun, Liu-Mei Zhang

Traditional similarity calculation method in collaborative filtering is inaccuracy due to the extreme sparsity of user rating data. To address this problem, we propose a collaborative filtering recommendation algorithm based on user integrated similarity. The algorithm modifies the similarity calculation formula by introducing the common factor. Then it introduces the item category interestingness eigenvector by category of items and distribution of user ratings to construct the user’s item category interestingness similarity. Finally, it combines the user rating similarity to construct the integrated similarity, and generates recommendations. The experimental results show that this algorithm can effectively relieve the inaccuracy of traditional similarity calculation method in the case of extreme sparsity of user rating data, and improve the quality of the recommendation of recommender systems.