The Construction And Application Of Knowledge Graph In The Field Of Coal Mine Accidents

With the rapid development of artificial intelligence in recent years,advanced technologies such as big data,Internet+,Internet of Things,cloud computing,etc.are widely used in the coal mine industry.At present,China's coal mine industry is increasingly moving toward intelligence and information technology.Although scientific and technological innovation has greatly promoted the development of the coal mine industry in recent years,there are still some problems in the field of coal mine accidents,such as complicated data information content,serious information fragmentation and single form.How to sort out the information,manage it and query it is an urgent problem to be solved.Therefore,the emerging field of knowledge graph is introduced into the field of coal mine accident,so that the data information in the field of coal mine accident can realize semantic connection through the knowledge graph,the multi-source heterogeneous data can be integrated and sorted out,and the data information and its connection can be vividly displayed by the visual graph,and accurate and fast information retrieval can be realized.This paper takes the knowledge of coal mine accident cases as the object,based on the deep learning method,studies the construction process and key technologies of the knowledge graph of coal mine accidents,and discusses the intelligent question answering query model and application of coal mine accidents based on the knowledge graph.First,according to the core set of coal mine accidents,the three types of knowledge entities,attributes and relationships in the coal mine accidents are established;secondly,the collected accident case information in the coal mines is serialized,and the current more commonly used BIOSE tagging specifications are used to identify the entity information.Make an annotation.Afterwards,through comparative experiments,the two-way LSTM model with the best training effect plus the conditional random field combination model was selected from the four algorithm models to complete the knowledge extraction part of the content,and the graph database Neo4j was used to store the knowledge graph to form the final coal mine Knowledge graph of accident domain;finally,based on the knowledge graph of coal mine accident domain,an intelligent question answering model is established for knowledge query.The model first sets the corresponding question types according to the feature words in the coal mine accident information and collects question sentences as training samples,uses jieba word segmentation,regular expressions and other techniques to extract relevant entity information in the question sentences,and then uses Naive Bayes classification The algorithm trains the question classification model,and then maps the key entity information of the question to the Cypher template to form a complete sentence.Run in the Neo4j database to search for the answer,and finally establish a window page that interacts with the user,allowing the user to select the question type and input Question sentence,click the search button,retrieve the answer to the question,and finally get a complete intelligent question and answer model in the field of coal mine accidents.This article organically integrates the field of coal mine accidents with the knowledge graph,combs and integrates the various and messy coal mine accident information to facilitate unified management,and also enriches the display form of coal mine accident information.Intelligent Q&A based on knowledge graph facilitates the inquiry and understanding of accident information,provide rich case data for coal mine safety management in the future,and lay a good foundation for subsequent research.Figure[31]Table[13]Reference[94]...