AIGC背景下科技学术生态风险的协同防范体系构建  

Construction of Collaborative Prevention System of Scientific and Technological Academic Ecological Risks under the Background of AIGC

作　　者：许立群 XU Liqun(Periodical Press,Jingchu University of Technology,Jingmen,Hubei 448000,China)

机构地区：[1]荆楚理工学院期刊社,湖北荆门448000

出　　处：《荆楚理工学院学报》2024年第5期81-90,共10页Journal of Jingchu University of Technology

摘　　要：人工智能生成内容(AIGC)背景下,学术生态风险被放大,包括基于真实研究的科技论文置信度降低、学术不端的科技论文数量激增、作者责任无人承担。而原有的风险防范手段仅能较好识别第三项风险,对于前两项风险识别效果欠佳,而这两项风险都会影响科技论文置信度。鉴于置信度不足的论文通常缺乏足够的科研事实支撑,提出把科技论文的审核和科研行为的真实性、科研结果的真实性结合起来的监管手段,构建协同防范体系。该体系包含长效的科研大数据比对机制、实验生成类图像溯源机制、对科研单位的监督机制、标准化的重复实验失败举报处理流程,使科技论文刊发的学术规范更系统更完备,有利于提高研究可靠性、保障科技论文置信度,并有助于提升学术风气,提高中国的国际学术话语权。In the context of artificial intelligence generated content(AIGC),academic ecological risks are amplified,including the risk of decreased credibility of scientific papers based on real research,the risk of a surge in the number of scientific papers based on academic misconduct,and the risk that the author is not held accountable.However,the original risk prevention method can only identify the third risk well,while the identification effect of the first two risks which will both affect the credibility of scientific and technological papers is poor.In view of the fact that papers with low credibility usually lack sufficient support of scientific research facts,we propose to combine the review of scientific papers with the authenticity of scientific research behavior and results to build a collaborative prevention system.The system includes a long-term big data comparison mechanism for scientific research,an image traceability mechanism for experimental generation,a supervision mechanism for scientific research institutions,and a standardized process for reporting and handling repeated experimental failures,which makes the academic norms of scientific and technological papers more systematic and complete,and is conducive to improving research reliability and guaranteeing the credibility of scientific and technological papers.It also helps to enhance the academic atmosphere and improve China's international academic discourse power.

关 键 词：生成式人工智能 人工智能生成内容 科技论文 学术生态 学术监管 科研管理 

分 类 号：G311[文化科学] G230[经济管理—管理学] C931.2