工业设备维修知识图谱构建与应用
Construction and Application of Industrial Equipment Maintenance Knowledge Graph

在工业智能化革命快速发展的助推之下，知识图谱正越来越多地应用到工业领域中。本文以某散热膜生产工厂生产设备维修记录为数据源，利用自顶向下和自底向上相结合的构建方式，首先对原始数据进行处理与分析，在专业人员的指导下，完成对知识图谱模式层(本体)的构建，再通过对数据源的抽取，按照模式层的本体间关系，构建知识图谱三元组；最后，通过Neo4j图数据库将其储存起来，并以节点和边的图谱方式直观地将所构建的工业设备维修知识图谱表示出来。本文所构建的工业设备维修知识图谱可应用于维修人员评价系统、智能问答系统，智能推荐系统以及维修异常预警系统。
Driven by the rapid development of industrial intelligent revolution, knowledge graph is increasingly applied to the industrial field. In this paper, we take the maintenance records of production equipment in a heat dissipation film production factory as the data source, and we use the construction method of the combination of top-down and bottom-up. Firstly, we process and analyze the original data, and complete the construction of the knowledge graph pattern layer (ontology) with the guidance of professionals. Secondly, through the extraction of data sources, according to the relationship between ontology in the pattern layer we construct the triples of knowledge graph; finally, the knowledge graph we construct is stored in the graph database called Neo4j. The constructed industrial equipment maintenance knowledge graph is intuitively expressed in the form of nodes and edges. This knowledge graph can be applied to build the evaluation system of maintenance personnel, intelligent question and answer system, intelligent recommendation system and maintenance anomaly early warning system.