Scholarlink

The process of applying for science and technology projects often encounters issues such as duplicate or redundant submissions, leading to a significant waste of resources. Traditional duplicate detection techniques based on text matching or vector similarity calculations struggle to effectively address semantic duplication in scientific documents. To meet the requirements of duplicate detection for science and technology project applications, we propose the development of a duplicate detection system grounded in an engineering knowledge graph. This system integrates the construction of an engineering knowledge graph, the establishment of a semantic retrieval workflow, and the development of a domain-specific large model for hydropower engineering. By leveraging the structural characteristics of scientific documents, the system performs comprehensive semantic duplicate detection across multiple dimensions, including titles, abstracts, keywords, paragraphs, and full texts. This approach enables the efficient identification of similar or related documents, thereby supporting enterprises in their project proposal and approval processes for science and technology initiatives.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Development of Document Plagiarism Detection System for Hydropower Engineering Based on Engineering Knowledge Graph

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Chao Zhang,
  • Zichang Li,
  • Peng Lin,
  • Yao Xu,
  • Jian Yang,
  • Fan Feng

摘要

The process of applying for science and technology projects often encounters issues such as duplicate or redundant submissions, leading to a significant waste of resources. Traditional duplicate detection techniques based on text matching or vector similarity calculations struggle to effectively address semantic duplication in scientific documents. To meet the requirements of duplicate detection for science and technology project applications, we propose the development of a duplicate detection system grounded in an engineering knowledge graph. This system integrates the construction of an engineering knowledge graph, the establishment of a semantic retrieval workflow, and the development of a domain-specific large model for hydropower engineering. By leveraging the structural characteristics of scientific documents, the system performs comprehensive semantic duplicate detection across multiple dimensions, including titles, abstracts, keywords, paragraphs, and full texts. This approach enables the efficient identification of similar or related documents, thereby supporting enterprises in their project proposal and approval processes for science and technology initiatives.