Slope One算法在Hadoop平台的改进实现

In recent years, E-commerce sites have show great interests on Precise Advertising. Collaborative filtering recommendation algorithm have been widely used as it is a solution to find users' exactly neeed. In this paper, research has shown that Slope One as one of collaborative filtering algorithm has been widely used for the sake of it is simple and easy to be implemented. But Slope One algorithm is also have a greate problem that it is difficult to deal with large intermediate data when training set is big. This study explore the c of combine of cloud computing and Slope One algorithm,.Then proposes a improved program implementation of Slope One algorithm on a open source cloud computing platform Hadoop. Based on the deep research of characteristics of Hadoop and Slope One algorithms, offer a description of the specific process to achieve this improvement, and then test this program with GropLens database ,at last give a conclusions of this solution based on the result.