Documentation, Informaiton & Knowledge ›› 2025, Vol. 42 ›› Issue (3): 44-52,117.doi: 10.13366/j.dik.2025.03.044

• Academic Focus: Large Language Models and Information Resources Management • Previous Articles     Next Articles

KELLM-Driven Intelligent Service of Scientific and Technological Insights: Connotation, Implementation, and Challenges

ZHOU Fanqian1, MAO Jin1,2, LI Gang1,2   

  1. 1. School of Information Management, Wuhan University, Wuhan, 430072;
    2. Center for Studies of Information Resources, Wuhan University, Wuhan, 430072
  • Online:2025-05-10 Published:2025-06-23
  • Contact: Correspondence should be addressed to MAO Jin, Email: danveno@163.com, ORCID: 0000-0001-9572-6709
  • Supported by:
    This is an outcome of the project "Research on Associative Scientific Hypotheses Generation Incutting Edge Interdisciplinary Areas by Incorporating Heterogeneous Knowledge Representations"(72474159)supported by National Natural Science Foundation of China, and the project "Research on Intelligent Service of Scientific and Technological Insights Based on Knowledge-Enhanced Large Models"(250025013)supported by the Open Foundation of Intelligence Engineering Laboratory, Institute of Scientific and Technical Information of China.

Abstract: [Purpose/Significance] In response to the intelligent challenges posed by technological innovation to intelligence insights in the digital intelligence era, this research systematically reviews the connotations, implementation pathways, and development challenges of intelligent service of scientific and technological insights, aiming to enhance the capabilities of intelligent scientific and technological intelligence services from a dynamic and holistic perspective. [Design/Methodology] Focusing on the scenario of scientific and technological intelligence, based on perceptual thinking, DIKW theory, and other related frameworks, this research explores the connotation of intelligent service of intelligence insights. It elucidates the theoretical logic of knowledge-enhanced large language models (KELLM)by combing the core capabilities and workflows of the model to empower intelligent intelligence insights, and proposes an implementation path driven by KELLM. Finally, the paper summarizes the challenges and opportunities faced by the intelligent service of scientific and technological insights in the current era. [Findings/Conclusion] The epoch-making technological leap, represented by artificial intelligence, will inevitably give rise to the intelligent service of scientific and technological insights. Knowledge-enhanced large language models will serve as a technical engine to empower the intelligent service of scientific and technological insights with these capabilities in knowledge association and integration, knowledge processing and analysis, intelligence comprehension and generation, as well as intelligence application optimization.[Originality/Value] This research proposes the concept of intelligent service of scientific and technological insights and constructs the implementation path based on knowledge-enhanced large language models, which can effectively guide the technology research and product development for the intelligent service of scientific and technological insights.

Keywords: Intelligence insight, Large Language Model, Knowledge enhancement, Artificial intelligence, Intelligence service