摘要
[目的/意义]作为中美竞争背景下美国政府制定对华政策的重要信息源,美国智库涉华成果的情报价值不言而喻。以数据驱动的研究范式系统性地分析和组织美国智库涉华成果,以求达成提升国内智库成果产出质量,增强政府部门决策科学性的目的。[方法/过程]在大情报观理论的指导下,立足于政府部门感知美国智库涉华政策的需求,梳理国外智库成果感知工作的特性以及其与图书情报研究中文献挖掘方法的共性,设计美国智库涉华内容分析框架,实现细粒度知识挖掘、全文本计量、信息检索等方法与智库成果分析工作的融合。在实证研究中基于多种自然语言处理技术挖掘智库涉华文献中的关键信息,构建美国智库涉华文献知识库与检索系统,并以该系统定量分析美国智库对南海问题的政策提议。[结果/结论]研究结果表明,提出的框架能够良好地适配于美国智库涉华文献的挖掘和组织,相关成果能为研究者提供感知国外智库政策的全局性的视角,有助于推动国内智库研究实现由理论驱动到数据驱动的转变。
[Purpose/Significance]As an important source of information to formulate US policy towards China within the context of Sino-US competition,the intelligence value of the China-related outputs of US think tanks is self-evident.This study systematically analyzes and organizes the outputs of American think tanks concerning China in a data-driven research paradigm,aiming to improve the quality of domestic think tank outputs and enhance the scientific nature of government decision-making.[Method/Process]Under the guidance of the big intelligence view,and based on the government departments’need to perceive American think tanks policies concerning China,this research outlined the characteristics of foreign think tanks output perception and its commonalities with literature mining methods in library and information science studies,and designed an analysis framework of U.S.think tanks content related to China.This framework integrated fine-grained knowledge mining,full-text measurement,information retrieval with think tank output analysis.In the empirical study,various natural language processing technologies were used to mine key information from China-related think tank literature,constructed a knowledge base and retrieval system,and quantitatively analyzed the policy proposals of American think tanks on the South China Sea issue.[Result/Conclusion]The results show that the proposed framework adapts well to the mining and organization of American think tank literature on China.The related outputs can provide researchers with a global perspective on perceiving foreign think tank policies and contribute to the transition of domestic think tank research from a theory-driven to a data-driven approach.
作者
刘畅
沈思
王东波
何洪旭
林立涛
赵志枭
Liu Chang;Shen Si;Wang Dongbo;He Hongxu;Lin Litao;Zhao Zhixiao(College of Information Management,Nanjing Agricultural University,Nanjing 210095;School of Economics&Management,Nanjing University of Science and Technology,Nanjing 210094;China Center for International Economic Exchanges,Beijing 100050;School of Information Management,Nanjing University,Nanjing 210023)
出处
《图书情报工作》
北大核心
2024年第12期77-94,共18页
Library and Information Service
基金
国家自然科学基金面上项目“基于深度学习的学术全文本知识图谱构建及检索研究”(项目编号:71974094)研究成果之一。
关键词
大情报观
美国智库
自然语言处理
知识组织
全文本计量
big intelligence
view American think tanks natural language processing knowledge organization full-text bibliometric
