Abstract: Coal mine safety laboratory has more than 30 existing experimental projects,the experimen⁃ tal project involves prediction of coal seam gas gushing, gas drainage technology, and prevention and control of coal and gas outburst. The existing laboratory information management system has accumulated mas⁃ sive experimental data. In order to explore the valuable information hidden behind the data at a deeper level, the data mining algorithm is applied, the laboratory information management system collectes a large number of data as the original data for preprocessing.The K-means clustering algorithm is used to establish the clustering model of the original data properties, and the fuzzy C-means algorithm is used for optimization and improvement. The results show that the evaluation results are consistent with the reality. The two data mining methods can effectively analyze the coal and gas outburst risk, and assist the coal and gas disas⁃ ter prevention. The application of data mining algorithm in the laboratory information management system can provide effective technical support for gas disaster prevention and control.

Key words: data mining,clustering,K-means algorithm,disaster prevention