Abstract Nowadays, the organizational structure of public service facilities is loose, which is not conducive to the development and progress of public services. Therefore, it is very important to find an optimization algorithm for the supply of public services in intelligent communities. Aiming at the resource waste problem of high repetition rate and low utilization rate in the current method, an optimization algorithm for providing public services in intelligent communities under the background of big data is proposed. According to the attributes of planning factors, the planning scheme of the spatial layout is divided into multiple subsystems and the goodness value of the factors in the scheme is calculated by using the single factor modular root method. Taking each subsystem as a factor, the comprehensive goodness value of each decision scheme is obtained respectively, so as to complete the creation of the spatial layout model of public service facilities. Based on this model, the fractal dimension, shape, and discrete index of a land patch are calculated for the public service facilities of the intelligent community, and the calculated results are used to optimize the public service supply in intelligent communities. The experimental results show that the proposed algorithm can optimize the public service supply of intelligent communities and give them the advantages of low repetition rate and high utilization rate.