GPT-3.5 Turbo和GPT-4 Turbo在系统评价的标题和摘要筛选中的应用

GPT-3.5 Turbo and GPT-4 Turbo in Title and Abstract Screening for Systematic Reviews.

作者信息

Oami Takehiko, Okada Yohei, Nakada Taka-Aki

机构信息

Department of Emergency and Critical Care Medicine, Chiba University Graduate School of Medicine, 1-8-1 Inohana, Chuo, Chiba, 260-8677, Japan, 81 432262372.

Department of Preventive Services, Kyoto University Graduate School of Medicine, Kyoto, Japan.

出版信息

JMIR Med Inform. 2025 Mar 12;13:e64682. doi: 10.2196/64682.

DOI:10.2196/64682

PMID:40073422

原文链接:https://pmc.ncbi.nlm.nih.gov/articles/PMC11922487/

Abstract

This study demonstrated that while GPT-4 Turbo had superior specificity when compared to GPT-3.5 Turbo (0.98 vs 0.51), as well as comparable sensitivity (0.85 vs 0.83), GPT-3.5 Turbo processed 100 studies faster (0.9 min vs 1.6 min) in citation screening for systematic reviews, suggesting that GPT-4 Turbo may be more suitable due to its higher specificity and highlighting the potential of large language models in optimizing literature selection.

摘要

这项研究表明，虽然与GPT-3.5 Turbo相比，GPT-4 Turbo具有更高的特异性（分别为0.98和0.51），以及相当的敏感性（分别为0.85和0.83），但在系统评价的文献筛选中，GPT-3.5 Turbo处理100篇文献的速度更快（分别为0.9分钟和1.6分钟），这表明GPT-4 Turbo可能因其更高的特异性而更适用，并突出了大语言模型在优化文献选择方面的潜力。