人工智能在撰写科学评论文章中的应用。

IF 4.3 2区 医学
Current Osteoporosis Reports Pub Date : 2024-02-01 Epub Date: 2024-01-16 DOI:10.1007/s11914-023-00852-0
Melissa A Kacena, Lilian I Plotkin, Jill C Fehrenbacher
{"title":"人工智能在撰写科学评论文章中的应用。","authors":"Melissa A Kacena, Lilian I Plotkin, Jill C Fehrenbacher","doi":"10.1007/s11914-023-00852-0","DOIUrl":null,"url":null,"abstract":"<p><strong>Purpose of review: </strong>With the recent explosion in the use of artificial intelligence (AI) and specifically ChatGPT, we sought to determine whether ChatGPT could be used to assist in writing credible, peer-reviewed, scientific review articles. We also sought to assess, in a scientific study, the advantages and limitations of using ChatGPT for this purpose. To accomplish this, 3 topics of importance in musculoskeletal research were selected: (1) the intersection of Alzheimer's disease and bone; (2) the neural regulation of fracture healing; and (3) COVID-19 and musculoskeletal health. For each of these topics, 3 approaches to write manuscript drafts were undertaken: (1) human only; (2) ChatGPT only (AI-only); and (3) combination approach of #1 and #2 (AI-assisted). Articles were extensively fact checked and edited to ensure scientific quality, resulting in final manuscripts that were significantly different from the original drafts. Numerous parameters were measured throughout the process to quantitate advantages and disadvantages of approaches.</p><p><strong>Recent findings: </strong>Overall, use of AI decreased the time spent to write the review article, but required more extensive fact checking. With the AI-only approach, up to 70% of the references cited were found to be inaccurate. Interestingly, the AI-assisted approach resulted in the highest similarity indices suggesting a higher likelihood of plagiarism. Finally, although the technology is rapidly changing, at the time of study, ChatGPT 4.0 had a cutoff date of September 2021 rendering identification of recent articles impossible. Therefore, all literature published past the cutoff date was manually provided to ChatGPT, rendering approaches #2 and #3 identical for contemporary citations. As a result, for the COVID-19 and musculoskeletal health topic, approach #2 was abandoned midstream due to the extensive overlap with approach #3. The main objective of this scientific study was to see whether AI could be used in a scientifically appropriate manner to improve the scientific writing process. Indeed, AI reduced the time for writing but had significant inaccuracies. The latter necessitates that AI cannot currently be used alone but could be used with careful oversight by humans to assist in writing scientific review articles.</p>","PeriodicalId":48750,"journal":{"name":"Current Osteoporosis Reports","volume":" ","pages":"115-121"},"PeriodicalIF":4.3000,"publicationDate":"2024-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10912250/pdf/","citationCount":"0","resultStr":"{\"title\":\"The Use of Artificial Intelligence in Writing Scientific Review Articles.\",\"authors\":\"Melissa A Kacena, Lilian I Plotkin, Jill C Fehrenbacher\",\"doi\":\"10.1007/s11914-023-00852-0\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p><strong>Purpose of review: </strong>With the recent explosion in the use of artificial intelligence (AI) and specifically ChatGPT, we sought to determine whether ChatGPT could be used to assist in writing credible, peer-reviewed, scientific review articles. We also sought to assess, in a scientific study, the advantages and limitations of using ChatGPT for this purpose. To accomplish this, 3 topics of importance in musculoskeletal research were selected: (1) the intersection of Alzheimer's disease and bone; (2) the neural regulation of fracture healing; and (3) COVID-19 and musculoskeletal health. For each of these topics, 3 approaches to write manuscript drafts were undertaken: (1) human only; (2) ChatGPT only (AI-only); and (3) combination approach of #1 and #2 (AI-assisted). Articles were extensively fact checked and edited to ensure scientific quality, resulting in final manuscripts that were significantly different from the original drafts. Numerous parameters were measured throughout the process to quantitate advantages and disadvantages of approaches.</p><p><strong>Recent findings: </strong>Overall, use of AI decreased the time spent to write the review article, but required more extensive fact checking. With the AI-only approach, up to 70% of the references cited were found to be inaccurate. Interestingly, the AI-assisted approach resulted in the highest similarity indices suggesting a higher likelihood of plagiarism. Finally, although the technology is rapidly changing, at the time of study, ChatGPT 4.0 had a cutoff date of September 2021 rendering identification of recent articles impossible. Therefore, all literature published past the cutoff date was manually provided to ChatGPT, rendering approaches #2 and #3 identical for contemporary citations. As a result, for the COVID-19 and musculoskeletal health topic, approach #2 was abandoned midstream due to the extensive overlap with approach #3. The main objective of this scientific study was to see whether AI could be used in a scientifically appropriate manner to improve the scientific writing process. Indeed, AI reduced the time for writing but had significant inaccuracies. The latter necessitates that AI cannot currently be used alone but could be used with careful oversight by humans to assist in writing scientific review articles.</p>\",\"PeriodicalId\":48750,\"journal\":{\"name\":\"Current Osteoporosis Reports\",\"volume\":\" \",\"pages\":\"115-121\"},\"PeriodicalIF\":4.3000,\"publicationDate\":\"2024-02-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10912250/pdf/\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Current Osteoporosis Reports\",\"FirstCategoryId\":\"3\",\"ListUrlMain\":\"https://doi.org/10.1007/s11914-023-00852-0\",\"RegionNum\":2,\"RegionCategory\":\"医学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"2024/1/16 0:00:00\",\"PubModel\":\"Epub\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Current Osteoporosis Reports","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1007/s11914-023-00852-0","RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2024/1/16 0:00:00","PubModel":"Epub","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0

摘要

综述的目的:最近,人工智能(AI),特别是 ChatGPT 的使用呈爆炸式增长,我们试图确定 ChatGPT 是否可用于协助撰写可信的、经同行评审的科学评论文章。我们还试图在一项科学研究中评估为此目的使用 ChatGPT 的优势和局限性。为此,我们选择了 3 个在肌肉骨骼研究中具有重要意义的主题:(1) 阿尔茨海默病与骨骼的交叉;(2) 骨折愈合的神经调控;(3) COVID-19 与肌肉骨骼健康。针对上述每个主题,我们采用了 3 种方法撰写稿件:(1) 仅人类;(2) 仅 ChatGPT(仅人工智能);(3) #1 和 #2 的组合方法(人工智能辅助)。为确保科学质量,对文章进行了广泛的事实检查和编辑,最终稿件与原稿有很大不同。在整个过程中测量了许多参数,以量化各种方法的优缺点:总体而言,使用人工智能减少了撰写综述文章的时间,但需要进行更广泛的事实核查。在仅使用人工智能的情况下,多达 70% 的参考文献被发现不准确。有趣的是,人工智能辅助方法的相似度指数最高,这表明剽窃的可能性更大。最后,尽管技术日新月异,但在研究期间,ChatGPT 4.0 的截止日期为 2021 年 9 月,因此无法识别近期的文章。因此,所有在截止日期之后发表的文献都是人工提供给 ChatGPT 的,这就使得第 2 和第 3 种方法对当代引文的识别完全相同。因此,对于 COVID-19 和肌肉骨骼健康主题,由于方法 2 与方法 3 有大量重叠,因此中途放弃了方法 2。这项科学研究的主要目的是了解是否能以科学适当的方式使用人工智能来改进科学写作过程。事实上,人工智能缩短了写作时间,但却存在严重的不准确性。后者要求目前不能单独使用人工智能,但可以在人类的认真监督下使用人工智能来协助撰写科学评论文章。
本文章由计算机程序翻译,如有差异,请以英文原文为准。

The Use of Artificial Intelligence in Writing Scientific Review Articles.

The Use of Artificial Intelligence in Writing Scientific Review Articles.

Purpose of review: With the recent explosion in the use of artificial intelligence (AI) and specifically ChatGPT, we sought to determine whether ChatGPT could be used to assist in writing credible, peer-reviewed, scientific review articles. We also sought to assess, in a scientific study, the advantages and limitations of using ChatGPT for this purpose. To accomplish this, 3 topics of importance in musculoskeletal research were selected: (1) the intersection of Alzheimer's disease and bone; (2) the neural regulation of fracture healing; and (3) COVID-19 and musculoskeletal health. For each of these topics, 3 approaches to write manuscript drafts were undertaken: (1) human only; (2) ChatGPT only (AI-only); and (3) combination approach of #1 and #2 (AI-assisted). Articles were extensively fact checked and edited to ensure scientific quality, resulting in final manuscripts that were significantly different from the original drafts. Numerous parameters were measured throughout the process to quantitate advantages and disadvantages of approaches.

Recent findings: Overall, use of AI decreased the time spent to write the review article, but required more extensive fact checking. With the AI-only approach, up to 70% of the references cited were found to be inaccurate. Interestingly, the AI-assisted approach resulted in the highest similarity indices suggesting a higher likelihood of plagiarism. Finally, although the technology is rapidly changing, at the time of study, ChatGPT 4.0 had a cutoff date of September 2021 rendering identification of recent articles impossible. Therefore, all literature published past the cutoff date was manually provided to ChatGPT, rendering approaches #2 and #3 identical for contemporary citations. As a result, for the COVID-19 and musculoskeletal health topic, approach #2 was abandoned midstream due to the extensive overlap with approach #3. The main objective of this scientific study was to see whether AI could be used in a scientifically appropriate manner to improve the scientific writing process. Indeed, AI reduced the time for writing but had significant inaccuracies. The latter necessitates that AI cannot currently be used alone but could be used with careful oversight by humans to assist in writing scientific review articles.

求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
Current Osteoporosis Reports
Current Osteoporosis Reports ENDOCRINOLOGY & METABOLISM-
CiteScore
8.40
自引率
2.30%
发文量
44
期刊介绍: This journal intends to provide clear, insightful, balanced contributions by international experts that review the most important, recently published clinical findings related to the diagnosis, treatment, management, and prevention of osteoporosis. We accomplish this aim by appointing international authorities to serve as Section Editors in key subject areas, such as current and future therapeutics, epidemiology and pathophysiology, and evaluation and management. Section Editors, in turn, select topics for which leading experts contribute comprehensive review articles that emphasize new developments and recently published papers of major importance, highlighted by annotated reference lists. An international Editorial Board reviews the annual table of contents, suggests articles of special interest to their country/region, and ensures that topics are current and include emerging research. Commentaries from well-known figures in the field are also provided.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信