Design And Implementation Of Intelligent Question Answering System For Railway Electrical Accidents Based On Knowledge Graph

With the rapid development of information technology,Cloud Computing,Artificial Intelligence,Internet of Things and Mobile Communications have been increasingly mature and widely used in various fields.Although the popularization of information technology promotes the rapid development of railway electrical accident field,there are still some problems in the field,such as wide distribution of data information and complex relationship.How to manage and maintain the data information in the field of railway communication and accurately query the relevant knowledge is still an urgent problem to be solved.Knowledge graph and intelligent question answering technology can solve the problems effectively.Therefore,the intelligent question answering system based on knowledge graph is innovatively introduced into the field of railway electrical accidents,which improves the maintainability of data and the convenience and accuracy of information retrieval.By visualizing and managing various data information of railway electrical accidents,the accuracy and efficiency of information query is optimized.In this paper,the construction of knowledge graph,the principle and method of intelligent question and answer is researched,and the design and implementation of intelligent question and answer system for railway communication is put forward.Knowledge graph,knowledge question answering,graph database,ECharts framework and entity recognition model are briefly introduced.Firstly,through the analysis of the knowledge characteristics of the railway electrical accident,the entity in the field,knowledge attributes and the relationships between knowledge are established.Therefore,the model design of the railway electrical accident knowledge is completed.Then,the initial data is annotated and the Bi LSTM-CRF neural network and regular expression are used to extract knowledge entities,entity attributes from the annotated corpus.In addition,the relationships between entities as well as entities and attributes are manually increased.Finally,the graph database Neo4 j is used to complete the construction and storage of the knowledge graph.A natural language intelligent question answering system is designed on the basis of this knowledge graph.Firstly,the system defines different question types according to the characteristic words in user’s questions and collects examples according to it.Entities in natural language questions are extracted by using Question Cleaning,Part-of-Speech Tagging and Name Entity Recognition techniques.Secondly,referring to the feature words of the input question,the Naive Bayes algorithm is applied to predict the types of user’s questions.Finally,the entity information in the user’s question is mapped to the Cypher sentence template corresponding to the question type,and the final Cypher sentence is generated.As a result,the answer is searched out from the Neo4 j database and returned to the user.In this research,an intelligent question and answer system based on knowledge graph is designed and implemented.The functions of user rights management,knowledge management,knowledge query and knowledge display of railway electrical accidents are realized.The feasibility and practicability of the system are verified through examples,and useful explorations are made for the construction of knowledge graphs and intelligent applications in the field of railway electrical accidents.