Abstract: The development of intelligence and informationization is the inevitable trend of coal enterprise safety production, and data mining technology plays an important role in promoting the development of coal mine safety management. Natural language processing technology and Bi-LSTM model are used to automatically classify coal mine hidden danger information and extract coal mine hidden danger factors; gas explosion risk is taken as the starting point to construct a gas safety posture prediction model; and an accurate control information system based on data mining technology and intelligent technology is established to realize the information analysis of big data and intelligent information processing at the application end. The system is applied to Yushupo Coal Mine in Shanxi Province, and the results show that it significantly improves the capacity and efficiency of coal mine safety management, and it provides new ideas and methods for the construction of dual pre-control in coal mines, enhances the information processing capacity of the existing pre-control system.