Easy Recommendation Based on Probability Model
Jian Chen, Jin Yan Huang, Huaqing Min · 2008
Traditional collaborative filtering recommendation system suffers from some significant limitations, such as scalability and sparsity, which cause the speed and quality of recommendation system is unacceptable. To alleviate these problems, this paper proposes a novel algorithm based on probability model. Our algorithm can directly generate the preference prediction from database and at the same time has the ability to reflect the changes of users' interest incrementally. The effectiveness of the new algorithm is estimated by our experiments.