基于数据决策分析的科研成果关注度影响因素研究  

Influencing Factors of the Attention of Scientific Achievements Based on Data Decision

作　　者：王正威 李海林[1,2] 陈多 万校基[1] WANG Zheng-wei;LI Hai-lin;CHEN Duo;WAN Xiao-ji(College of Business Administration,Huaqiao University,Quanzhou 362021,China;Research Center of Applied Statistics and Big Data,Huaqiao University,Quanzhou 362021,China)

机构地区：[1]华侨大学工商管理学院,福建泉州362021 [2]华侨大学现代应用统计与大数据研究中心,福建泉州361021

出　　处：《情报科学》2022年第7期27-36,共10页Information Science

基　　金：国家自然科学基金项目“高维时间序列数据聚类分析及应用研究”(71771094);福建省社会科学规划项目“基于文献主题时间序列数据挖掘的技术预见研究”(FJ2020B088)。

摘　　要：【目的/意义】科研成果关注度是论文价值成果的体现,反映研究者对成果的关注程度,也影响科研成果的转化效率,科研人员可依据关注度确定研究方向与主题。【方法/过程】提出科研成果关注度概念。量化潜在影响科研成果关注度的因素,包括关键词、作者地址、基金资助、期刊热度、标点符号吸引力和作者水平。【结果/结论】以影响因素为条件属性,关注度为决策属性,使用CART决策树获得决策规则,研究发现:(1)商业与经济学研究方向中化学、生物、自然等领域的成果更易受关注;(2)关键词对科研成果关注度具有高度影响,其次为期刊热度,而部分因素也对科研成果关注度存在不同程度的影响;(3)基金资助和标题符号吸引力在独立分析中无法证实与科研成果关注度的关系,但在决策过程发现二者没有影响。【创新/局限】本文基于学者的关注度视角,提出了科研成果关注度概念,用机器学习模型进行决策,为进一步评估科研成果提供了理论基础。【Purpose/significance】The attention of scientific achievements is the embodiment of the value of one article,which reflects the researchers’attention for the article,and affects the efficiency of transformation of scientific achievements,meanwhile researchers can determine research directions and topics according to attention.【Method/process】This paper proposes the concepts of the attention of scientific achievements.Potential influencing factors are quantified including keywords,author’s address,funding,heat of journals,punctuation attraction and author level.【Result/conclusion】The influence factor is the conditional attribute,and attention is the decision attribute,the study found that:(1)In business and economics research area,achievements of chemistry,biology,nature get more attention.(2)Keywords have a high impact on attention,the second is the heat of journals,and sectional factors have different degrees of influence on attention.(3)The relationship between funding,punctuation attraction and the attention of scientific achievements could not be verified in independent analysis,but they have no impact on it in the process of decision classification.【Innovation/limitation】Based on the perspective of scholars’attention,this paper proposes concepts related to the attention of scientific achievements.Machine learning model is used to make decisions,which provides a theoretical basis for further evaluation of scientific achievements.

关 键 词：科研成果 关注度 影响因素 近邻传播聚类 决策规则 

分 类 号：G250.2[文化科学—图书馆学]