Construction of Vocational Education Knowledge Graph Based on DeepSeek

In the context of the digital and intelligent era, exploring effective paths for the digital transformation of vocational education through artificial intelligence technology is of vital importance. This research takes the “Compressor” chapter of the “Civil Aircraft Maintenance Personnel License Management Rules” (CCAR-66-R3) license course “Gas Turbine Engine” as the object and proposes an automatic construction method for vocational education course knowledge graphs based on the DeepSeek large language model (LLM). By designing targeted prompt words (Prompts), the model API is invoked to extract (entity-relation-entity) triples from unstructured textbook texts, and Neo4j graph database is used for storage and visualization. Eventually, a structured knowledge network representing the structure, working principle, and fault diagnosis of the compressor system is formed. This method verifies the feasibility of large language models in the structured processing of complex technical field knowledge and provides a reference technical solution for the efficient construction of knowledge bases on the vocational education field and the support of intelligent teaching applications.