In this paper, a platform for digital leisure sports services is built using digital information technology in the context of the new economy. Plain Bayesian classification is used to classify sports information based on its features, and the target value is assigned based on the feature vector. The RBF algorithm is introduced to transform the information into measurable data, and the target weights are updated using the energy function. The results indicate that the villagers exercised for 5 hours each day on the platform, and the BMI pass rate of villagers of different age groups exceeded 70%. The digital innovation of leisure sports has a greater effect on the psychology and physiology of villagers and is a strong fit for the prosperity and revitalization of rural culture.