LLM-powered framework enhances STI analytics with dynamic knowledge graphs

This paper introduces a novel framework that combines large language models (LLMs) with dynamic knowledge graphs and traditional bibliometrics to enhance the analysis of science, technology, and innovation (STI). The proposed system addresses the limitations of static bibliometric indicators by integrating LLMs for semantic enrichment, but strictly constrains their use to provisional candidate generation. These candidates undergo a multi-layer validation process, ensuring epistemic discipline and reproducibility before being used for advanced analytics like trend emergence and pathway mapping. AI

IMPACT This framework could lead to more accurate and timely insights into scientific and technological advancements, aiding research and policy decisions.

RANK_REASON The item is an academic paper proposing a new framework for STI analytics. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• alphaXiv
• bibliometrics
• CatalyzeX Code Finder for Papers
• Connected Papers
• DagsHub
• Gotit.pub
• Hugging Face
• knowledge graph
• large language models
• Litmaps
• ScienceCast
• scite Smart Citations

• paper
• infra

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →