Multidimensional impact of large language models on scientific research: Methodological integration, efficiency gains, and shifts in funding patterns

Zhang, Yiqin; Zhu, Yu; Gong, Hongcun; Wang, Fanming; Deng, Sanhong

Permalink

https://hdl.handle.net/2142/126210 Copy

Description

Title

Multidimensional impact of large language models on scientific research: Methodological integration, efficiency gains, and shifts in funding patterns

Author(s)

• Zhang, Yiqin
• Zhu, Yu
• Gong, Hongcun
• Wang, Fanming
• Deng, Sanhong

Issue Date

2025-03-11

Keyword(s)

• Large Language Models
• Scientific Research Methodologies
• Interdisciplinary Collaboration
• AI Integration

Abstract

Introduction. The rapid evolution of Large Language Models (LLMs) has significantly transformed scientific research. This study aims to explore the multidimensional impact of LLM integration, focusing on its influence on research methodologies, efficiency, and collaboration across various disciplines. Method. A systematic content analysis was conducted using a comprehensive dataset sourced from Web of Science. Indicators such as LLM integration levels, use cases, shifts in research methodologies, and impacts on funding and research outcomes were analyzed. The dataset was manually curated to include 722 relevant academic articles, ensuring rigorous data selection. Analysis. The analysis focused on tracking the progression of LLM integration in scientific research from 2019 to 2024. Specific attention was given to identifying trends in LLM applications for data analysis, hypothesis generation, and interdisciplinary collaboration, as well as methodological transformations. Results. The findings indicate a substantial increase in LLM integration, particularly in 2023, where their use in core research activities peaked. LLMs have significantly enhanced computational methods and fostered interdisciplinary collaboration, although some challenges, such as overreliance and ethical concerns, remain. Conclusion(s). LLM technologies are poised to revolutionize scientific research by improving efficiency, creativity, and collaboration. Future research should focus on addressing the ethical implications of LLM use and further exploring their long-term impact on scientific methodologies and innovation.

Publisher

iSchools

Series/Report Name or Number

iConference 2025 Proceedings

Type of Resource

Other

Genre of Resource

Conference Poster

Language

eng

Handle URL

https://hdl.handle.net/2142/126210

Copyright and License Information

Copyright 2025 is held by Yiqin Zhang, Yu Zhu, Hongcun Gong, Fanming Wang And Sanhong Deng. Copyright permissions, when appropriate, must be obtained directly from the authors.

Owning Collections