Research on fault Q&A system for train-controlled on-board equipment based on knowledge graph

At present, the presentation of the knowledge graph of train-controlled on-board equipment is only one model in most cases, the entity extraction process is more complicated, and there is a lack of Q&A system and other related practical applications. In view of the above problems, the paper proposes to apply the deep learning model to the entity recognition in order to improve the accuracy rate, F₁, and increase the faults of the other four models to make the presentation of the graph more complete, and ultimately combine the knowledge graph, graph database, natural language processing, knowledge quiz and other technologies are combined with this field. This paper adopts entity extraction,relationship extraction, entity alignment and knowledge storage to create a knowledge map of train-controlled on-board equipment faults, and accordingly realises the construction of a Q&A system to provide auxiliary decision-making for railway maintenance personnel.