Related Experiment Video
Updated: Jan 16, 2026
03:14
Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
Published on: December 6, 2024
1.0K
The Landscape of Wearable Sensors and Automated Literature Analysis with Large-Language Models
Jaromir Klarák1, Vitor H B D Santi2, Luan F Moreira2
1Institute of Informatics, Slovak Academy of Sciences, Bratislava 845 07, Slovak Republic.
ACS Omega
|September 29, 2025
Summary
This study combines network analysis and large language models (LLMs) to automate literature reviews on wearable sensors for health monitoring. While promising for scalable analysis, LLMs require human oversight for accuracy and depth.
Area of Science:
- Biomedical Engineering
- Computer Science
- Materials Science
Background:
- Scientific literature growth necessitates efficient analysis tools.
- Wearable sensors are crucial for health monitoring.
- Automated literature synthesis is an emerging research area.
Purpose of the Study:
- To explore integrating complex network analysis and LLMs for automated literature reviews.
- To focus on the research field of wearable sensors for health monitoring.
- To assess the effectiveness and limitations of this hybrid approach.
Main Methods:
- Utilized OpenAlex for retrieving scientific papers on wearable sensors.
- Constructed paper citation networks and performed thematic clustering.
- Employed LLMs (ChatGPT) to generate summaries from network clusters.
Main Results:
- Identified key subtopics like graphene sensors, gait analysis, and AI applications.
- LLMs generated coherent overviews but showed superficiality and inaccuracies.
- Demonstrated potential for scalable literature reviews with hybrid methods.
Conclusions:
- Combining network analysis and LLMs offers a framework for accelerating scholarly synthesis.
- Human oversight is essential for ensuring the rigor and relevance of automated analyses.
- Identified critical gaps in wearable sensor technology, including standardization and energy efficiency.
