Abstract: To effectively reduce coal mine accidents caused by unsafe acts, it is necessary to construct the identification and role analysis model based on FTA and 24 Model. Through empirical research, the 6-step application process was concretized and 22 causes of unsafe actions were identified, involving 30 action paths. The results show that the FTA identification program can effectively display the logical relationship among various factors, and UA-UA and UA-UC involve a total of 20 paths, with a strong correlation. Gephi0.9.2 classification features visualization highlights the correlation between the classification of actions, categories of personnel and specific violations, with violations accounting for 54.6% of the total number of violations, a total of 25 articles of the regulations.