Study On The Construction Of Coal Mine Safety Integration Knowledge Map

With the constantly deepening process of intelligent coalmines,users can search massive coalmine safety data,acquire relevant knowledge of coalmine safety,and catch the development trend of coalmine industry on the web.But these web data is large-scale,heterogeneous,diverse and lowly organized,which makes it difficult for users to identify effective information quickly and results in poor utilization of data and confusion of knowledge management.Thus,a knowledge base that can integrate,sort out,store and manage coalmine safety information is particularly important.Therefore,the knowledge graph is selected as the knowledge management method to transform the coalmine safety knowledge into a graph with high accuracy,wide coverage and clear logic,simplify the manual data selection process,and effectively improve the level of coalmine safety production.This paper mainly focuses on the construction of fusion knowledge graph of coalmine safety,puts forward the semi-automatic construction method of coalmine safety concept knowledge base,designs the automatic construction process of coalmine disaster event database,and internalizes the event database into the concept knowledge base in an orderly and hierarchical way,so as to preliminarily explore the integration of coalmine safety knowledge graph into disaster events.The main work of this paper is as follows:1)Semi-automatic construction of coalmine safety concept knowledge base.Firstly,we extract the key entities of the text,build the key entity database,propose the fusion-feature textrank keyword extraction algorithm based on optimized word-graph,improve the traditional textrank algorithm only rely on the word co-occurrence statistical characteristics to build the word graph weight matrix,integrate the text semantic and structural information.Then,clustering analysis is carried out on the entity database to form a general knowledge system,and the defect that it is difficult to initialize the parameters Eps and MinPts by optimizing the traditional DBSCAN algorithm is proposed based on the improved DBSCAN algorithm to complete entity clustering.Finally,combined with the expert knowledge to design the logical relationship between entities,the coalmine safety concept knowledge base is constructed semiautomatically,and the corresponding event nodes are added in each knowledge category to make a foundation for the integration of the subsequent event base.2)Automatic construction of coalmine disaster event database.Combined withthe coalmine disaster scene,a new method of building disaster event database is proposed.Firstly,the event detection is completed based on HIDCNN to detect whether the text belongs to the event text and its corresponding event category.Then,combined with the named entity recognition algorithm based on DMHSA-CRF and dependency syntax analysis,the extraction template of elements is designed to complete the task of extracting disaster elements in this paper,and the disaster event database is constructed.Finally,the disaster event database is added to the corresponding event node of the coalmine concept knowledge base,and the coalmine safety fusion knowledge graph is constructed,and stored in the Neo4 j graph database according to the pre-designed storage strategy.The paper has 47 figures,29 tables,78 references.