Abstract: Coal industry exhibits a notably high tendency for accident occurrences due to the inherently high-risk nature of its operations. To address the core demand of safeguarding workers’ occupational safety and health, there is an urgent need to leverage intelligent technologies for the accurate identification of the causes of potential safety risks in coal mine production. First, feature extraction and dimensionality reduction of hidden danger information were carried out on relevant textual data such as coal mine accident investigation reports. By adopting a coupled research method combining text mining and complex networks, 189 key features of safety accidents were extracted from a wide range of unstructured coal mine accident investigation reports, and 9-dimensional feature clusters were identified using the agglomerative hierarchical clustering method. Second, correlation tests were applied to determine the correlation between accident-causing features and calculate the correlation intensity among them, thus constructing a safety accident causation network. Finally, key hidden danger nodes were analyzed by combining the complex network model, and a data-driven early-warning strategy for safety accidents was proposed, which provides a reference for more efficient hidden danger detection and remediation as well as for improving the level of safe production. The results indicate that the key accident causes include factors such as “working face”, “driving face”, “goaf”, “steel wire rope”, and “conveyor”. The “working face” holds a central position in the characteristic network of coal mine safety accident hazards and exhibits a high degree of correlation with both “goaf” and “shearer”. These elements frequently co-occur during the evolution of accidents. The coupled model of text mining and complex networks can effectively analyze the chain propagation path of coal mine accident causes, and the data-driven early-warning strategy can improve the efficiency of hidden danger detection, provide assistance for the preventive control and management of coal mine safety risks.