With the development and popularization of the sharing economy in cities, the city's appearance and lifestyle have changed with each passing day, and the service functions of some old communities can no longer meet the current people's living needs. It is of practical significance for the sustainable and coordinated development of the city to study the strategies and methods of optimizing the functional space of the old community, optimize the cultural space of the old urban community and establish the shared service function. Aiming at the spatial information demand in urban development decision-making, this paper studies the theory, method and application of urban spatial dynamic data mining, and applies particle swarm optimization (PSO) to the optimization of cultural spatial layout of old urban communities with performance constraints, and discusses the method and application of urban land use spatial structure prediction. The example results verify the feasibility and effectiveness of the algorithm. The optimization of cultural space in old urban communities combined with PSO algorithm can obtain reasonable, feasible and scientific spatial optimization results, and the operation efficiency is obviously improved. It is particularly important to formulate and improve the laws and regulations of community development, formulate scientific and reasonable community planning, and attach importance to the inheritance and innovation of community culture. Using PSO can efficiently complete the layout optimization design of community cultural space.