Abstract: Mine hoist is the key equipment in mine safety production which is a large and complex system of machine, electric and hydraulic pressure. Because of the complexity of its own structure and the uncertain factor in operation conditions, the faults may be multiple failures, associated faults or other complex forms. So, it is essential to study on the effective technology of fault diagnosis for the mine hoist. Based on the analysis of traditional fault diagnosis method, this study proposed a new one which is in semantic environment. First of all, it constructed the fault diagnosis system structure based on ontology and described the function of each part, the next; it analyzed and established the mine hoist fault ontology. Once more, the mapping is achieved from fault phenomenon ontology to the fault cause one through the new association rules algorithm of affairs reduction, it solved the matching problem of fault ontology and provided a strong theoretical basis to hoist fault diagnosis. So the fault source is located and the most suitable fault solution is found out. In the end, the result is tested by experiment. Experiment results show that the method can solve the problem of knowledge sharing and reuse in fault diagnosis system and accurately and effectively realize the fault diagnosis for hoist and provide a reasonable proposal for diagnostic decision.