Abstract: Despite the rapid development of research and application of smart mines, coal mines’ current processes such as fully mechanized mining equipment fault diagnosis, disaster emergency rescue plan generation, disaster accident cause analysis, and production organization and operation decision-making still mainly rely on artificial experience and are not highly intelligent. An important reason for the above problems is that the current technical approach to smart mines mainly revolves around data, algorithms and computing power. Without effective use of domain expert knowledge, it is difficult to achieve autonomous decision-making. Facing the high-level construction goal of autonomous decision-making in smart mines, it is urgent to carry out research on the construction of knowledge graphs and reasoning methods in the coal mine field to form a four-element technical support system of “knowledge + data + algorithm + computing power”. Firstly, this article reviews the current research status of knowledge graphs, especially knowledge graphs in the coal mining field, and outline the development history of knowledge-driven artificial intelligence, the artificial intelligence system architecture based on knowledge graphs, the main types and representative work of knowledge graphs, and analyze the knowledge modeling situation, knowledge graph construction methods, knowledge graph usage methods and maturity of existing knowledge graphs in the coal mining field. Secondly, the challenges faced by knowledge graph construction and reasoning technology in the coal mine field are analyzed, covering aspects such as entity recognition, relationship extraction, knowledge graph fusion and error correction, and knowledge graph reasoning. Finally, the technical trends and application scenarios of knowledge graph construction and reasoning in the coal mining field are prospected. Through sorting and analysis, the following conclusions are drawn: (1)The existing knowledge graph construction goals in the coal mining field are relatively limited, the technical methods are relatively traditional, and it is difficult to support complex applications of intelligent decision-making; (2) The knowledge graph construction and reasoning technology in the coal mine field faces many challenges, including the difficulty of entity identification caused by the widespread presence of nested entities, the difficulty of relationship extraction caused by overlapping entities, the difficulty of entity alignment caused by the heterogeneity of knowledge graphs, the difficulty of correcting errors in knowledge graphs due to unclear consistency constraints on relationships between entities, and in combining knowledge graph reasoning technology with business scenarios; (3) The future application prospects of knowledge graphs in the coal mine field are broad and the demand is urgent. Therefore, it is imperative to do in-depth research on the construction and reasoning methods of knowledge graphs in this field.