Modeling of intelligent auxiliary decision-making for power system under multi-source heterogeneous data scenario
Accurate retrieval of power system information plays an important role in guiding the whole life cycle management of power systems.Due to the existence of data silos and low accu-racy of answer search,,power systems are in urgent need of new strategies to improve data stor-age and extraction efficiency.To address the above problems,this paper proposes a power system for multi-source heterogeneous data scenarios,mainly with the help of domain knowledge graph construction technology and related knowledge question and answer methods to intelligently as-sist subsequent business decisions.Firstly,the existing multi-source heterogeneous power system data are merged,and the concepts and axioms in the power system are defined in an ontology modeling way.Then an unsupervised Chinese entity-relationship extraction model based on the dependency semantic paradigm is used to extract triples from the fine-grained entities and rela-tionships in the Chinese power domain text to construct a knowledge graph,store the domain knowledge graph with the help of the Neo4j graph database,and implement fine-grained power system domain Q&A based on domain question template matching.The effectiveness of the pro-posed method is then verified by an interactive Q&A prototype system.
power system managementknowledge graphontology modelingquestion answer-ing system
NETL
NSTL
万方数据
