The public service system of national fitness is the condition and guarantee for the public to participate in various fitness activities. In order to effectively integrate the existing national fitness service resources and to improve the efficiency of the public service supply, this paper aims to study the informatization of building national fitness public services in smart city under the background of big data. Firstly, we build a public service information platform of national fitness in the smart city, consisting of different modules, and then describe each module in detail. Secondly, we consider different types of fitness projects and establish the database structure model of network resources. Thirdly, we obtain the multitree cascade system of the national fitness and the heterogeneous data model of the national fitness benefit index by using high-dimensional statistical analysis method. Combined with the national fitness behavior, data mining, and support vector machine (SVM), the objective function of data mining statistical decision-making is established. Finally, the well-known particle swarm optimization (PSO) method is used to optimize the target parameters so as to realize the national fitness, big data mining, and feature analysis. The simulation results show that the proposed model has superior performance, as it can realize heterogeneous data mining of the national fitness benefit index. In addition, we quantitatively analyze the promoting effect of the national fitness on the physical quality and health level of the public.