As the Korean government agencies have recently announced the first official road map for urban air mobility (UAM) with the aim of introducing a new aviation transportation system, many companies in South Korea have initiated projects to design airspace infrastructure for UAM operations at the early stage of development. Given that the agencies tentatively plan to expand UAM to regional air mobility (RAM) operations at the mature stage of development, this research specifically focuses on establishing airspace infrastructure for upcoming RAM operations in South Korea. The proposed methodology leverages three different algorithms: 1) a partitioning-based clustering algorithm for placing vertiport locations, 2) a density-based clustering algorithm for predicting areas of convective weather, and 3) the Latin hypercube sampling-based probabilistic road map (LHS-based PRM) algorithm for generating an adaptive airspace network. The resulting airspace, constructed by the proposed methodology, takes into account airspace restriction areas such as prohibited areas or military operation areas. The main contribution of this research is to employ a data-driven approach using machine learning and LHS-based PRM algorithms to dynamically establish airspace infrastructure to be potentially used for upcoming RAM operations in South Korea.