图书情报知识 ›› 2017, Vol. 0 ›› Issue (3): 118-129.doi: 10.13366/j.dik.2017.03.118

• 知识、学习与管理 • 上一篇    

知识图谱在科技情报研究中的应用模型构建

焦晓静,王兰成,韩锋   

  • 出版日期:2017-05-10 发布日期:2017-05-10

Application Model Construction of Mapping Knowledge Domains in the Science and Technology Intelligence Research

  • Online:2017-05-10 Published:2017-05-10

摘要:

当前,网络化和大数据的出现对科技情报研究构成了严峻挑战,对以定性分析为主的传统科技情报研究方法进行革新已迫在眉睫,知识图谱通过对大样本学术信息(如期刊论文、专利报告等)的分析来对科学知识的结构、关系与演化过程进行可视化,将其引入科技情报研究之中,有助于科技情报研究解决现存问题和应对现有挑战。本文概述了“事实型数据+专用方法工具+专家智慧”的国内科技情报研究通用模式以及美国罗伯特〖DK〗·克拉克提出的“以目标为中心”的情报研究模式,针对知识图谱的功能及其局限性、结合科技情报研究的特点,提出了知识图谱在情报研究中的三个应用策略,即:多源数据融合、分析方法的协同运用和科学解读图谱结果,并以知识图谱的应用场景为切入点,建议在科技情报研究四个环节的工作中引入知识图谱:情报研究选题、事实型数据库构建、情报分析、专家群体遴选。在此基础上,设计了知识图谱方法在科技情报研究中应用的优化模型,并提出了知识图谱应用流程。模型实现了定量与定性方法、情报人员与学科专家、知识图谱与其他情报研究方法等多种集成,为今后利用知识图谱方法开展科技情报研究奠定了基础。

关键词: 知识图谱, 科技情报研究, 信息可视化, 数据挖掘, 文献计量分析, 引文分析, 共现分析

Abstract:

The arrival of cyberization and big data era has posed a serious challenge to science and technology intelligence research. It is imperative to innovate the traditional intelligence research methods which give priority to qualitative analysis. Mapping knowledge domains is an information visualization method, which can visualize the structure, relationships and evolutionary process of scientific knowledge based on the analysis to the large sample of academic information (such as journal articles, patent report).In this paper, the general pattern of domestic science and technology intelligence research of “factual database+ special tools+ the wisdom of experts” and the “targetcentric” intelligence analysis process put forth by Robert M. Clark are overviewed. Based on the analysis to the function and boundedness of mapping knowledge domains and to the characteristics of science and technology intelligence research, three application strategies are put forward:multisource data fusion, collaborative use of analysis methods and scientific interpretation of mapping results. Putting the application scenarios of mapping knowledge domains as the breakthrough point, it is suggested that mapping knowledge domains could be applied to four intelligence research processes:topic selection, construction of factual database, intelligence analysis and expert group selection. Based on the discussion above, the optimization model of mapping knowledge domains’ application in science and technology intelligence research is designed. Then the application flow of mapping knowledge domains is put forward. Several integrations are realized through the application model-the integration of quantitative analysis and qualitative analysis, the integration of intelligence personnel and subject experts, mapping knowledge domains and other intelligence research methods, etc. The model lays a foundation for the follow-up application of mapping knowledge domains in science and technology intelligence research.

Key words: Mapping knowledge domains, Science and technology intelligence research, Information visualization, Data mining, Bibliometric analysis, Citation analysis, Co-occurrence analysis