本文对知识图谱关键技术进行研究，提出一种多数据源融合的电力物资供应商评价知识图谱构建流程。采用自顶向下知识图谱进行本体构建，再进行知识抽取、知识表示、知识融合，最后进行知识存储。实现对不同信息系统内的供应商评价知识图谱，方便查询和可视化展示。实现从多个维度对供应商评价数据进行汇总，得到综合评价，为对供应商制定精准监造策略提供了坚实的数据依据。
In this paper, the key technology of knowledge graph is studied, and a process of building knowledge graph of power material supplier evaluation based on multi data source fusion is pro-posed. The top-down knowledge graph is used for ontology construction, knowledge extraction, knowledge representation, knowledge fusion, and finally knowledge storage to realize the knowledge graph of supplier evaluation in different information systems, which is convenient for query and visual display. It can summarize the supplier evaluation data from multiple dimensions and get comprehensive evaluation, which provides a solid data basis for the formulation of precise manufacturing supervision strategy for suppliers.
