摘要
本文利用文本挖掘技术,对山东省大学生创业政策进行了动态演进规律分析。借助2006—2022年国家、省、市三个层级的57份关于山东省大学生创业政策文本,利用共词分析、LDA主题建模与相似度计算三种技术分别进行文本挖掘,并利用CiteSpace与LDAvis方法进行可视化分析。研究发现,(1)政策关键词共现频率方面,存在“环境型最多、需求型其次、供给型最少”、“层级滞后性”等规律。(2)政策主题内容方面,“层级阶段关联性”特征较为明显。(3)政策工具使用方面,存在明显的“层级滞后发展性”、“中央政策作用强、地方政策作用弱”等特征。进一步地,本文从减少政策时间滞后性、增强政策一致性、完善政策内涵等方向提出建议。
This paper analyses the dynamic evolution law of college student entrepreneurship policy in Shandong Province using text mining technology.On the 57 texts on university student entrepreneurship policies in Shandong Province at the national,provincial and municipal levels from 2006 to 2022,text mining was carried out by using three techniques,namely,co-word analysis,LDA theme modelling and similarity calculation,and visual analysis was carried out by using CiteSpace and LDAvis methods.The following conclusions are obtained:(1)In terms of the co-occurrence frequency of policy keywords,there are the laws of"the most environmentally oriented,the second most demand oriented,and the least supply oriented"and"hierarchical lag".(2)In terms of the policy theme content,the feature of"hierarchical stage correlation"is more obvious.(3)The use of policy tools,there are obvious"hierarchical lag development","central policy role is strong,local policy role is weak"and other characteristics.Further,this paper puts forward suggestions from the direction of reducing policy time lag,enhancing policy consistency and improving policy content.
作者
于文成
于淼
YU Wen-cheng;YU Miao(School of Economics and Management(Cooperative College),Qingdao Agricultural University,Qingdao 266109,China)
出处
《青岛农业大学学报(社会科学版)》
2024年第2期50-58,62,共10页
Journal of Qingdao Agricultural University(Social Science)
基金
山东省教育科学“十四五”规划课题“基于身份转换的山东省大学生创业驱动因素与引导政策研究”(2021QYB009)。
关键词
大学生创业
政策
文本挖掘
可视化分析
student entrepreneurship
policy
text mining
visual analytics