Construction of Domain Knowledge Graph Base for Coal Mine Rock‐Burst Prediction and Prevention Based on Ontology

ABSTRACT In China, as the mining depth of coal mines increases, the occurrence conditions of coal seams are becoming increasingly complex, leading to a continuous rise in the risk of rock‐burst. The frequent occurrence of such accidents has caused severe losses to the safe production of coal mines. With the advancement of coal mine informatization, it is essential to construct an ontology‐based knowledge graph in this field to accurately predict and prevent rock‐burst and promote the sharing and reuse of relevant knowledge. On the basis of relevant research results on rock‐burst, this study clarified the scope of knowledge involved. After extracting relevant terms using the TextRank algorithm, it constructed concept classes, such as mine geology, mining technology, prediction and evaluation, accident manifestations, and prevention and control measures. These concept classes were sorted out and summarized to refine and extract corresponding subclasses. Object properties and data properties were determined to further build the knowledge graph. According to the application logic of the knowledge map, referring to the comprehensive index method, the relevant rules of coal mine rock‐burst and safety regulations, the reasoning rules of the semantic web rule language were constructed. Through these rules, the risk grade of rock‐burst in a coal mine was predicted, and relevant prevention measures and suggestions were given. The ontology‐based method for constructing the knowledge graph proposed in the study had realized knowledge‐based prediction of rock‐burst. The relevant prevention and control measures can assist in analyzing the causes of accidents, make disaster prevention decisions, and effectively improve the intelligent management level of coal mine safety production.