According to the characteristics of gas pipe-line field,the technical framework and multi-source data of knowledge graph in gas pipeline field are ana-lyzed.The Bi-LSTM-CRF model is selected as the knowledge extraction model to extract knowledge from data related to gas pipeline leakage,define the relation-ship types,clarify the relationship between entities,and use the Neo4j graph database to store and construct the knowledge graph of gas pipe-line leakage.By analyzing the information query method of gas pipeline leakage knowledge graph,pre-dicting and analyzing the evolution path of gas pipeline leakage,and inferring the treatment measures,the ef-fectiveness of the application of knowledge graph in gas pipeline field is verified.The next steps of applica-tion are prospected.
关键词
燃气管道泄漏/双向长短期记忆网络条件随机场模型/知识图谱
Key words
gas pipeline leakage/bidirectional long short-term memory network and conditional random field(Bi-LSTM-CRF)model/knowledge graph
维普
万方数据