With the advancement of Industry 4.0, computer terminals are increasingly integrated with the public Internet. While this integration improves operational efficiency and flexibility, it also brings new data management problems. Before starting the research, the theoretical knowledge of this research is defined and expressed. After that, the computer terminal data is obtained, and it is found that this data has problems such as redundancy and multidimensionality. In order to improve the efficiency of user data management, data extraction method based on PCA-ReliefF, data mapping method based on kernel principal components are designed and analyzed by numerical simulation using Matlab 7.1.The KPCA algorithm has a longer computing time than the PCA algorithm, which indicates that the kernel function is introduced on top of the original one in order to realize the nonlinear mapping from low-dimension to high-dimension, which results in the growth of the computing time. Although the computing time grows, the accuracy of KPCA is much higher than PCA, i.e., the introduction of kernel function in traditional PCA can improve the accuracy of computerized multidimensional data mapping, which facilitates the users to better manage the data of computer terminals.
