Abstract: The water inrush accident is the third largest coal mine accident after the gas accident and the roof accident. The analysis and exploration of the causes of the water inrush accident and the intrinsic relationship between the various factors can effectively realize the control and containment of the water inrush accident. The existing coal mine water inrush accident research mostly aims at a certain area or a certain aspect. There is a lack of in-depth research on the complex causal relationship among the influence factors and the influence degree of each factor on the accident. In order to solve this problem, decision making trial and evaluation laboratory (DEMATEL) and interpretative structural modeling method (ISM) are used to analyze the influencing factors of coal mine water inrush accidents. The multi-level hierarchical structure model is constructed, which is mapped into the Bayesian network (BN) model. The DEMATEL-ISM-BN model is obtained. Based on the data-driven theory, typical accident cases are studied. There are 18 influencing factors inducing coal mine water inrush accidents determined. Based on expert scoring results, DEMATEL analysis is carried out. The influence degree, influenced degree, cause degree and centrality of each factor are calculated. The reachability matrix of ISM is calculated according to the DEMATEL analysis results. The multi-level hierarchical structure model is constructed. The BN model is constructed based on the real case data of coal mine water inrush accidents. The causal chain analysis is carried out by using the fault diagnosis function of the BN model. The results of the DEMATEL analysis show that the main factors affecting the occurrence of coal mine water inrush accidents are the lack of understanding of water disasters and inadequate hydrogeological detection. The other factors include confusion of safety management and weak technical means. ISM analysis results show that the "three violations" behavior and water source threat are at the top of the multi-level hierarchical structure model of water inrush accidents. These are the direct factors inducing water inrush accidents. The BN analysis results show that the most likely cause chain is inadequate hydrogeological detection → water source threat → water inrush accidents. To effectively curb the occurrence of coal mine water inrush accidents, it is suggested to improve the staff awareness of water disasters, strictly carry out hydrogeological exploration, and fundamentally eliminate the illegal behavior of production personnel.