Personalized recommendation based on the improved similarity and fuzzy clustering
Zebin Wu, Yan Chen, Taoying Li · 2014
Recommendation system was one of the most important technologies of personalized service, and collaborative filtering technology had been successfully applied to personalized recommendation system. But traditional collaborative filtering algorithms existed sparsity, bad expansibility, real-time response speed and other defects, and those directly affected the quality of recommendation system. This paper proposed a personalized recommendation algorithm based on improved similarity and fuzzy clustering. Firstly, improving the similarity calculation method was used to solve the problem of similarity due to the inaccuracy caused by the sparseness. Then, the algorithm used user's similarity fuzzy clustering and search for the nearest neighbors of the target user on this basis, thus narrowing the scope of the nearest neighbors to find. Finally, the algorithm used the matrix after clustering to produce recommendations. The experimental results show that this method can effectively improve the accuracy of the recommendation system and alleviate data sparseness problem and improve the real-time response speed.