The emergency plan for construction accidents is an important part of emergency management, but traditional text-based plans have limitations in the efficiency and accuracy of information retrieval. In view of this, a construction and intelligent application method for emergency plans for construction engineering accidents based on knowledge graph technology was proposed. According to the principle of knowledge extraction, the structural characteristics of emergency plan texts were analyzed, and the entity and relationship joint extraction strategy was designed by integrating rule and language technology platform to extract key information. The Neo4j graph database was used to store and manage the knowledge graph. Case verification was carried out around the key links of emergency response. The results show that the constructed knowledge graph can quickly match emergency measures, support efficient query, statistics and intelligent question and answer, and help improve the utilization efficiency and intelligence level of emergency management knowledge.