Abstract: Taking gas disaster risk macro early warning of all coal mines within the jurisdiction of all levels of coal mine safety supervision departments and mining group companies, it is pointed out that the basic data of regional coal mine gas disaster risk prediction has the characteristics of multi-source, heterogeneous, massive, multi-dimensional, etc., and there are some problems in data collection, such as incomplete information collection and single data collection mode, solidified dimensions, etc.; the basic data of regional coal mine gas disater risk prediction is divided into four types： natural environment risk data of regional mine, production system risk data of regional mine, gas prevention risk data of regional mine, and macro-security environmental risk data; collection technologies of coal mine safety monitoring and gas disaster early warning data with structured characteristics, supervision and law enforcement inspection data with semi-structured characteristics, and coal mine audio and video monitoring data with unstructured characteristics are introduced, and the research of coal mine safety monitoring and gas disaster warning data collection technology based on .NET Core cross-platform Web API, and macro-security environmental risk data collection technology based on topic crawler are introduced in detail; data collection system of gas disaster risk early warning for regional coal mines suitable for Internet environment is designed, and field test show that the system can comprehensively, reliably and timely collect basic data of regional coal mine gas disaster risk early warning.