This work-in-progress mixed-methods study, set in the context of Danish research libraries, investigates the functionalities, perceptions, and adoption of generative AI (GenAI) tools that interact with scientific digital libraries to enhance the literature review process. The study focuses on identifying the names and functionalities of GenAI tools, as well as librarians’ perceptions and experiences in providing training and applying these tools for literature reviews. Data were collected through semi-structured interviews and a survey involving six librarians, desk research to map the functionalities of GenAI tools, and analysis of GenAI training content from a Danish university library. The open and axial coding methods were applied to analyze the interview data. Key barriers identified include lack of funding, technical expertise, data protection compliance, and trialability. Future research should focus on developing a literature review workflow journey map supported by GenAI tools, creating a GenAI evaluation instrument, and establishing an impact assessment framework.

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

Generative AI Tools for Systematic Approaches to Literature Review: Insights from Danish Librarians

  • Md Saifuddin Khalid,
  • Ali Marwan El-Haj Moussa,
  • Jiayan Wu,
  • Jeannette Ekstrøm

摘要

This work-in-progress mixed-methods study, set in the context of Danish research libraries, investigates the functionalities, perceptions, and adoption of generative AI (GenAI) tools that interact with scientific digital libraries to enhance the literature review process. The study focuses on identifying the names and functionalities of GenAI tools, as well as librarians’ perceptions and experiences in providing training and applying these tools for literature reviews. Data were collected through semi-structured interviews and a survey involving six librarians, desk research to map the functionalities of GenAI tools, and analysis of GenAI training content from a Danish university library. The open and axial coding methods were applied to analyze the interview data. Key barriers identified include lack of funding, technical expertise, data protection compliance, and trialability. Future research should focus on developing a literature review workflow journey map supported by GenAI tools, creating a GenAI evaluation instrument, and establishing an impact assessment framework.