Scholarlink

CNAF supports over 1500 users across 60 scientific communities through its user support department, handling queries related to hardware and software technologies. To face the increased demand for support expected in the upcoming future, a retrieval-augmented generation model is proposed, automating email support by using advanced natural language processing solutions. By leveraging a LangChain-powered framework and a custom retrieval-augmented generation architecture, a vector database from the Tier-1 User Guide has been integrated into a large language model. This approach enhances responsiveness and accuracy by utilizing specific domain knowledge and semantic similarity models to generate human-like responses.

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

Advanced NLP Solutions for Enhancing the CNAF User Support

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Alberto Trashaj,
  • Matteo Barbetti,
  • Elisabetta Ronchieri,
  • Carmelo Pellegrino,
  • Daniele Cesini,
  • Carmen Giugliano,
  • Daniele Lattanzio,
  • Lucia Morganti,
  • Alessandro Pascolini,
  • Andrea Rendina,
  • Aksieniia Shtimmerman

摘要

CNAF supports over 1500 users across 60 scientific communities through its user support department, handling queries related to hardware and software technologies. To face the increased demand for support expected in the upcoming future, a retrieval-augmented generation model is proposed, automating email support by using advanced natural language processing solutions. By leveraging a LangChain-powered framework and a custom retrieval-augmented generation architecture, a vector database from the Tier-1 User Guide has been integrated into a large language model. This approach enhances responsiveness and accuracy by utilizing specific domain knowledge and semantic similarity models to generate human-like responses.