Construction and Application of NLP-based Knowledge Graph in CNC Equipment Fault Field

Computer Numerical Control (CNC) equipment is a technology-intensive mechatronics complex system cov-ering multidisciplinary knowledge. How to effectively convert historical fault data into useful fault knowledge base is an urgent problem to be solved. A construction technology of CNC equipment fault knowledge graph based on Natural Language Processing (NLP) is proposed. BERT deep learning classification technology is used to build the sample classification model, the named entity recognition model is trained based on the Bi-LSTM technology, as well as the CNC equipment fault corpus sample data is trained, recognized, and modeled. Finally, Neo4j is used to form a knowledge graph model. It has been verified that the recognition rate of the model basically meets the requirements, and on this basis, a recommended solution for repairing equipment faults is proposed, which realizes the effective use of fault knowledge.