Efficient Training Corpus Retrieval for Large Language Model Fine Tuning: A Case Study in Cancer
Studies in Health Technology and Informatics · 2025
DOI: 10.3233/shti251039 ↗Kurz gefasst
Fachartikel in der Zeitschrift Studies in Health Technology and Informatics, 2025. Verfasst von Avisha Das und 6 weiteren Autor:innen. Beteiligte Einrichtung: Mayo Clinic Arizona, Phoenix, AZ, USA. Hauptthemen laut MEDLINE: Natural Language Processing, Urinary Bladder Neoplasms, Neoplasms, Data Mining. Die Arbeit wurde bisher 1-mal zitiert (Crossref). Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0). Gefördert von: NLM NIH HHS.
Abstract (Originalsprache)
The objective is to create an automated knowledge extraction tool for cancer research that builds high-quality academic corpora for LLM fine-tuning while investigating its effectiveness in interleukin-6 and bladder cancer domains. To address the current gap in knowledge retrieval techniques for cancer research data collection, we propose KnowledgePipeline, a novel automated tool that incorporates diverse aspects of academic papers and metadata. Our tool integrates content, co-citations, and co-authorship networks to construct domain-specific academic corpora suitable for fine-tuning LLMs. We leverage two LLMs (GPTJ-6.7B and Galactica30B) trained on domain-specific question-answer pairs from the refined data. The system’s evaluation focuses on both the quality of extracted knowledge and the performance of fine-tuned models in open-ended question-answering tasks. We see that KnowledgePipeline offers a scalable, automated framework for domain-specific knowledge retrieval and fine-tuned applications in cancer research, advancing literature discovery and addressing critical biomedical challenges. It achieved high relevance scores of 68% for IL-6 and 74.5% for bladder cancer, with a fine-tuned Galactica-30B model demonstrating promising capabilities.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Substanzen
Autor:innen und Institutionen
- Mayo Clinic Arizona, Phoenix, AZ, USA.
- University of Texas Health Science Center, Houston, TX, USA.
Förderung
- NLM NIH HHS
Zitieren
APA
Das, A., Diala, C., Chen, G., Li, Z., Li, R., Anjum, O., & Zheng, W. J. (2025). Efficient Training Corpus Retrieval for Large Language Model Fine Tuning: A Case Study in Cancer. Studies in Health Technology and Informatics. https://doi.org/10.3233/shti251039
Vancouver
Das A, Diala C, Chen G, Li Z, Li R, Anjum O, et al. Efficient Training Corpus Retrieval for Large Language Model Fine Tuning: A Case Study in Cancer. Studies in Health Technology and Informatics. 2025. doi:10.3233/shti251039
BibTeX
@article{Das2025,
author = {Das, Avisha and Diala, Chiamaka and Chen, Guocai and Li, Zhao and Li, Rongbin and Anjum, Omer and Zheng, W Jim},
title = {Efficient Training Corpus Retrieval for Large Language Model Fine Tuning: A Case Study in Cancer},
journal = {Studies in Health Technology and Informatics},
year = {2025},
doi = {10.3233/shti251039}
}
EUVIMED