| 袁庆,沈锡宾,刘红霞,王立磊,霍永丰,刘爽,陈晓峰,万贤贤,董文杰,刘冰,贺志阳,温文,周洪.中国科技期刊编辑大模型技术认知及其影响的调研研究.编辑学报,2024,(2):183-188 |
| 中国科技期刊编辑大模型技术认知及其影响的调研研究 |
| A survey research on the awareness and impact of large model technologies among Chinese STM journal editors |
| |
| DOI：10.16811/j.cnki.1001-4314.2024.02.014 |
| 中文关键词: 大模型技术 科技期刊 编辑 调查研究 |
| 英文关键词: |
| 基金项目:*中国科协大模型技术对中国科技期刊发展的影响分析及对策课题(2023KJQK012) |
| 作者 | 单位 | | 袁庆 | 《中华健康管理学杂志》编辑部
国家新闻出版署医学期刊知识挖掘与服务重点实验室,100052,北京 | | 沈锡宾 | 中华医学会杂志社新媒体部
国家新闻出版署医学期刊知识挖掘与服务重点实验室,100052,北京 | | 刘红霞 | 中华医学会杂志社新媒体部
国家新闻出版署医学期刊知识挖掘与服务重点实验室,100052,北京 | | 王立磊 | 中华医学会杂志社新媒体部
国家新闻出版署医学期刊知识挖掘与服务重点实验室,100052,北京 | | 霍永丰 | 《中华医学杂志》编辑部:100052,北京
国家新闻出版署医学期刊知识挖掘与服务重点实验室,100052,北京 | | 刘爽 | 《中华血液学杂志》编辑部,300052,天津 | | 陈晓峰 | 《科技进步与对策》编辑部,430071,武汉 | | 万贤贤 | 《科技进步与对策》编辑部,430071,武汉 | | 董文杰 | 《化学进展》编辑部,100190,北京 | | 刘冰 | 中华医学会杂志社,100052,北京
国家新闻出版署医学期刊知识挖掘与服务重点实验室,100052,北京 | | 贺志阳 | 讯飞医疗科技股份有限公司,230081,合肥 | | 温文 | 南京智齿数汇信息科技有限公司,210000,南京 | | 周洪 | 约翰威立国际出版集团Wiley partner solution 智能产品及人工智能研发部,100044,北京 |
|
| 摘要点击次数: 20 |
| 全文下载次数: 190 |
| 中文摘要: |
| 为全面了解中国科技期刊编辑对大模型技术的认知、应用现状及其他方面的影响,本研究向科技期刊编辑通过问卷星发放调研问卷,问卷包括国内外大模型技术和平台的认知情况、在期刊出版不同流程的应用情况,对大模型技术的需求、期望、风险和担忧,以及对大模型技术的投入意愿。本次问卷共回收450份有效问卷,82.44％的调研对象表示对大模型技术有过了解,但比较了解和非常了解仅占16.67％。对于大模型平台的认知方面,国外的除ChatGPT外,其他平台/工具认知度非常低；国内的以文心一言、星火大模型为多。在科研型平台方面,国内的Aminer为认识较多的平台。在实践应用阶段,目前在稿件组织和采编、内容传播和知识服务环节相对较好,在编辑和生产出版环节应用相对较少。74.89％的编辑对于大模型技术保持乐观和非常乐观的态度,仅不到2％的编辑表示悲观。在编辑的各环节中,认为相对比较容易被替代的工作环节是排版生产与数据加工,最不易被替换的是同行评议与选题策划,但文字编辑与传播运营方面也有超过60％的认为是可以被大部分替代的。为应对大模型技术的影响,编辑需要在技术应用、政策规范、前沿技术、出版伦理等多方面接受培训,尤其对技术应用方向,9成编辑需要得到相关的信息。编辑对于大模型技术抱有开放拥抱的态度,并亟须获得相关的培训与教育机会,以解决他们在应用场景上的困难,以及在政策法规及诚信建设方面的困惑。 |
| 英文摘要: |
| To comprehensively understand the awareness, current application, and other impacts of large model technologies on Chinese STM journal editors, aiming to provide necessary references for formulating related policies for Chinese STM journals,a survey was distributed to Chinese STM journal editors through “Wenjuanxing” questionnaire tool, covering the awareness of large model technologies and platforms both domestically and internationally, their application in different journal publishing processes, the demand, expectations, risks, and concerns regarding large model technologies, as well as the willingness to invest in them. A total of 450 valid questionnaires were collected, with 82.44% of respondents acknowledge large model technologies, but only 16.67% had a comprehensive or very comprehensive understanding. In terms of the awareness of large model platforms, the recognition of foreign platforms/tools, except for ChatGPT, was very low; domestically, ERNIE Bot and Spark Cognitive Toolkitwere more recognized, with Aminer being a well-recognized tool for scientific research. In the practical application stage, the use of large model technologies is relatively better in manuscript organization, content compilation, dissemination, and knowledge services, but less so in editing and production publishing. 74.89% of editors are optimistic or very optimistic about large model technologies, with less than 2% expressing pessimism. Among various editorial processes, they believe that type setting and data processing are relatively easy to be replaced, while peer review and topic planning are the least likely to be replaced, but more than 60% also believe that text editing and dissemination operations can be largely substituted. Chinese STM journal editors have a preliminary understanding of large model technologies, and some editors have already used tools to assist in their work.However, the overall understanding is not deep, and the application is not comprehensive. Editors have an open and embracing attitude towards large model technologies and urgently need relevant training and educational opportunities to solve their difficulties in application scenarios, as well as confusion in policies, regulations, and integrity construction. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
