Abstract: In view of the problem that the existing coal and gas outburst early warning technology does not make sufficient use of data analysis, this paper introduces the architecture, big data types and acquisition methods of the outburst early warning technology system based on big data from each technical link of the big data life cycle. A multi-parameter coal and gas outburst early warning indicator system has been established from three aspects of working face, area, and production system, and the early warning model of coal and gas outburst has been established based on AHP, at the same time, the calculation method of relative importance of each factor in AHP based on the multi-layer association rule algorithm is obtained, which realizes the automatic determination and optimization of index weight; the coal and gas outburst early warning system based on B/S architecture model is developed, which mainly includes early warning database, early warning service, early warning website and mobile terminal APP. The field test shows that the total accuracy of early warning reaches 91.5% and there is no underreporting, realizing the dynamic acquisition, intelligent analysis, real-time early warning, online release and remote query of early warning data. The application of big data technology to the prevention and control of coal and gas outburst disasters is the future development direction. It is feasible to carry out outburst early warning based on big data, which can provide decision support for the prevention and control of coal and gas outburst in coal mines.