A knowledge graph construction method for compliance review of water conservancy project reports
Xinhang Zhang, Tao Wang, Xin Du, Zhefei Fan, Yuanfeng Hao
PLoS ONE
Problems Identified (4)
Manual compliance review bottlenecks: Manual compliance review of water conservancy project reports has efficiency and accuracy bottlenecks.
Need for smart water conservancy digital transformation: The water conservancy domain needs technical support to promote digital transformation toward Smart Water Conservancy.
Manual compliance review bottlenecks: Manual compliance review of water conservancy project reports has efficiency and accuracy bottlenecks.
Need for smart water conservancy digital transformation: The water conservancy domain needs technical support to promote digital transformation toward Smart Water Conservancy.
Proposed Solutions (5)
Knowledge graph construction for compliance review: The paper proposes constructing a knowledge graph for compliance review of water conservancy project reports.
BERT-BiLSTM-CRF entity recognition: The method uses a BERT-BiLSTM-CRF model for named entity recognition to locate engineering parameters and normative clauses.
CFG relation parsing to triples: The method uses context-free grammar and semantic labels to parse logical relationships and transform normative clauses into head-relation-tail triples.
Neo4j graph storage and update: The method stores the knowledge graph in Neo4j and uses Py2neo for efficient import and dynamic updating of triple data.
Knowledge graph construction for compliance review: The paper proposes constructing a knowledge graph for compliance review of water conservancy project reports.
Results (3)
Feasibility verified in case review:
Improved review efficiency and accuracy:
Technical support for digital transformation:
Research Domain
Water conservancy project compliance review / knowledge graph construction