Xiamen Sharelink Technology Co.,LTD
The fast development of Internet has led the user into the information age. Large amount of information has been presented in the internet. Moreover, the amount of information is increasing extremely quickly and human is unable to process it. Recommendation system has been widely used in electronic commerce, news, web2.0, E-learning and other fields. Collaborative filtering is one of the most important algorithms. But as scale of recommendation system continues to expand, more and more problems appear. Data sparsity and poor prediction are main problems that recommendation system has to face. To improve the quality and performance, a new collaborative filtering recommendation algorithm combining user-clustering and Slope One algorithm is proposed.