Large Language Models in Scientific Research

The field of scientific research is witnessing a significant shift towards the adoption of large language models (LLMs) to improve various aspects of research, including topic discovery, ontology learning, and metadata harvesting. Researchers are leveraging LLMs to develop innovative methods for generating research topic ontologies, enhancing topic discovery in scientific literature, and creating scalable scientific interest profiling systems. These advancements have the potential to accelerate ecological research, facilitate easier scientific information retrieval, and enable researchers to gain deeper insights into emerging trends. Notable papers in this area include:

  • Scalable Scientific Interest Profiling Using Large Language Models, which presents a system for generating scientific interest profiles using LLMs, demonstrating moderate semantic similarity with self-written profiles.
  • Heterogeneous LLM Methods for Ontology Learning, which showcases a comprehensive system for addressing the full ontology construction pipeline, achieving top-ranking results in the official leaderboard across all three tasks.
  • SciTopic: Enhancing Topic Discovery in Scientific Literature through Advanced LLM, which proposes an advanced topic discovery method enhanced by LLMs, outperforming state-of-the-art scientific topic discovery methods.
  • Leveraging Large Language Models for Generating Research Topic Ontologies: A Multi-Disciplinary Study, which investigates the capability of LLMs to identify semantic relationships among research topics, demonstrating excellent performance across all disciplines.

Sources

Scalable Scientific Interest Profiling Using Large Language Models

Annif at the GermEval-2025 LLMs4Subjects Task: Traditional XMTC Augmented by Efficient LLMs

Estimating the Effective Topics of Articles and journals Abstract Using LDA And K-Means Clustering Algorithm

Heterogeneous LLM Methods for Ontology Learning (Few-Shot Prompting, Ensemble Typing, and Attention-Based Taxonomies)

Flexible metadata harvesting for ecology using large language models

SciTopic: Enhancing Topic Discovery in Scientific Literature through Advanced LLM

Leveraging Large Language Models for Generating Research Topic Ontologies: A Multi-Disciplinary Study

Built with on top of