Research on Abstracts Extraction Method of Power Accident Cases
As the complexity of power system increases,abstract extraction of power accident cases is of great significance for accident analysis and prevention.In this paper,starting from the structural,temporal and spatial features of power accident cases,we adopt text mining technology to extract keywords,utilize natural language processing technology for entity and event extraction,and construct accident descriptions through semantic role labeling and information fusion.The coherence and completeness of the abstracts are enhanced by combining knowledge graph and logical reasoning,and finally the abstracts are automatically generated through templates and text generation models,with a view to improving the extraction efficiency and quality of the abstracts of electric power accident cases,and providing a scientific basis for electric power safety management.
power accidentsabstract extractionnatural language processingknowledge graphevent extraction
万方数据
维普
