Abstract: To address the challenges of high index dimensionality and the difficulty of a single weighting method in balancing expert experience with the statistical laws of measured data in rockburst risk assessment under intelligent mining conditions,this study proposes a visual identification method for rockburst hazard in intelligent working faces based on Bayesian combination weighting. From the three dimensions of static load,medium properties,and dynamic load disturbance, an index system comprising 22 indicators across five categories—including burial depth and geological structure—is constructed through the analysis of rockburst disaster-causing mechanisms,expert consultation,and on -site investigation. The Analytic Hierarchy Process (AHP) is employed to determine the subjective weights. Principal Component Analysis (PCA) is introduced to eliminate weakly correlated indicators, such as redundant features, thereby reducing the evaluation dimensionality. Based on Bayesian theory,a relative entropy objective function is established,and the objective weights derived from information entropy are taken as the likelihood function to revise the subjective weights,thereby achieving a nonlinear coupling of subjective and objective weights. Furthermore, combined with cloud model theory, the comprehensive evaluation model is divided into four modules:a mechanism module,a feature reconstruction module,a Bayesian optimization module,and a cloud map visualization module. The compilation and encapsulation of the script programs for each module are completed, enabling intelligent optimization of weight-index correlations and visual identification of rockburst hazard levels. Sample analysis demonstrates that the model first identifies and eliminates weakly correlated indicators such as uniaxial compressive strength,and recognizes the distance to faults as a strongly coupled risk source of geological structure and mining disturbance. Meanwhile, the weight of indicators such as residual coal thickness are significantly revised,which effectively captures dynamic risks and provides data support for the practical prevention and control of mine rockburst and production decision-making. The final case study yields a Grade I:no hazard evaluation for a working face in Ordos,which is highly consistent with the on-site results obtained by the comprehensive index method, thereby verifying the accuracy and confidence of the proposed method in rockburst hazard identification.