In order to achieve intelligent analysis and treatment of rock bursts in coal mines, as well as facilitate the sharing and reuse of knowledge related to this phenomenon, this paper proposes the construction of a knowledge base for predicting and preventing mine rock bursts based on ontology. Firstly, following an analysis of the requirements for ontology construction and a comparison of existing methods, we employed a seven-step approach to develop the ontology specific to rock bursts. Utilizing the Textrank algorithm, we extracted significant entities pertinent to the prediction and prevention of rock bursts from relevant literature as sources for conceptual terms. Subsequently, we analyzed the interrelationships among these types before constructing a comprehensive knowledge base on rock bursts in coal mining using ProTéGé software. The rules were formulated based on a composite index method alongside SWRL language. Taking a coal mine as an illustrative example, we deduced the risk level associated with rock bursts according to established rules and proposed appropriate preventive measures. This paper advocates for establishing a knowledge base aimed at enhancing predictions and preventative strategies regarding rock bursts through ontological frameworks. Furthermore, it aims to enable rule-based reasoning concerning such predictions while assisting in analyzing causes behind impact accidents and informing decision-making processes related to disaster prevention and control measures—ultimately improving safety intelligence within coal mining operations.