摘要
【目的/意义】以科研人员在生成式人工智能应用中知识采纳行为作为研究对象,系统、全面识别其关键影响因素并厘清因素间的层级关系和关联路径。【方法/过程】通过元人种志结合质性访谈方法析取出科研人员生成式人工智能应用中知识采纳行为影响因素,基于全面质量管理理论将22个影响因素分为个体、技术、知识、组织和社会五个维度,使用模糊DEMATEL方法识别出11个关键影响因素,并运用ISM模型对影响因素进行层级划分和关联路径分析。【结果/结论】研究发现,在11个关键影响因素中,知识匹配度是影响科研人员在生成式人工智能应用中知识采纳行为的根源因素,需求和动机、对新技术的接受程度和知识内容是影响科研人员在生成式人工智能应用中知识采纳行为的最直接因素。【创新/局限】文中以科研人员为研究对象,从五个维度出发研究其生成式人工智能应用中知识采纳行为关键影响因素,具有一定的创新性。但影响因素提取时具有一定的主观性,未来可进一步量化影响因素,具体测算各个影响因素的作用大小。
【Purpose/significance】This study takes the knowledge adoption behavior of researchers in the application of generative artificial intelligence as the research object,systematically and comprehensively identifies the key influencing factors and clears the hierarchical relationship and correlation path among the factors.【Method/process】Through meta-ethnography combined with qualitative interview method,factors influencing knowledge adoption behavior of researchers in the application of generative artificial intelligence were analyzed.Based on total quality management theory,22 influencing factors were divided into five dimensions:individual,technology,knowledge,organization and society,and 11 key influencing factors were identified using fuzzy DEMATEL method.The ISM model is used to divide the influencing factors and analyze the correlation path.【Result/conclusion】It is found that among the 11 key factors,knowledge matching is the root factor that affects the knowledge adoption behavior of researchers in the application of generative artificial intelligence,and demand and motivation,acceptance degree of new technology and knowledge content are the most direct factors that affect the knowledge adoption behavior of researchers in the application of generative artificial intelligence.【Innovation/limitation】This paper takes researchers as the research object and studies the key influencing factors of knowledge adoption behavior in the application of generative artificial intelligence from five dimensions,which is innovative to a certain extent.But it is inevitably subjective when extracting the influencing factors.In the future,the influencing factors can be further quantified and the effects of each influencing factor can be measured concretely.
作者
郭顺利
张雪宁
GUO Shunli;ZHANG Xuening(Department of Communication,Qufu Normal University,Rizhao 276826,China)
出处
《情报科学》
北大核心
2025年第5期58-69,共12页
Information Science
基金
国家社会科学基金青年项目“基于认知计算的网络问答社区知识的深度聚合及精准服务研究”(20CTQ028)。
