The invention discloses a mineralization system source area detection method and system based on big data mining, and a medium, and relates to the technical field of geological mineral exploration. Gravity data, magnetic data, electrical method data, artificial seismic data and natural seismic data obtained by detecting a target research area are obtained; feature extraction, correlation analysis, clustering analysis and classification processing are performed on the gravity data, the magnetic data, the electrical method data, the artificial seismic data and the natural seismic data in sequence to obtain a source area detection result of the target research area, so that a lithosphere structure, possible earth dynamics events and metallogenic combination and distribution of the target research area can be determined; according to the method, the problems of prospecting where and what mine is found are solved based on the geodynamics events and the lithosphere structure, the efficiency, comprehensiveness and accuracy of source area detection of the metallogenic system are improved, prospecting exploration and target area optimization are supported, and the exploration risk is reduced.