Online preschool education relies heavily on resource creation, however incorrect construction location is a major issue. The issue of network resource building in preschool education network teaching cannot be solved by typical neural network algorithms, and the results are unsatisfactory. Consequently, this study reviews previous work on the topic of building online teaching materials for preschool education and suggests new research on building preschool education networks based on the gray clustering method. In order to minimize interference elements in resource building, resource development. After that, a data mining strategy is developed using the gray clustering method, and the outcomes of this strategy. The demonstrate that the gray clustering algorithm outperforms the classic neural network approach when it comes to resource construction affecting factor time and accuracy.