This research aims to optimize the resource utilization of metallurgical solid waste in a data-driven manner. By systematically constructing a large database and integrating machine learning models, we seek to optimize resource utilization pathways and establish an intelligent predictive system, thereby supporting the national strategic goal of green and low-carbon sustainable development.