Research on System of Data Mining Technology Based on Computer

Abstract

The existing implicit feedback collaborative algorithm directly uses sparse binary social trust information to assist recommendation, which has serious data sparse problems and does not have the influence of deep integration of social trust information. In response to the above problems, the paper proposes an algorithm that uses a denoising autoencoder to deeply integrate user implicit feedback data and social information. At the same time, a cloud computing-based social trust information data mining system architecture is constructed, and its modules are described. This framework is conducive to solving the problems of high data complexity and large amount of data encountered in current data mining, and can significantly improve the performance of data mining.