图书情报工作 ›› 2023, Vol. 67 ›› Issue (2): 108-118.DOI: 10.13266/j.issn.0252-3116.2023.02.011

• 知识组织 • 上一篇    下一篇

基于知识图谱的循证政策中科学证据推荐研究——以新冠肺炎疫情防控政策为例

任超1, 杨孟辉1,2, 杨冠灿1, 霍朝光1, 卢小宾1   

  1. 1 中国人民大学信息资源管理学院 北京 100872;
    2 数据工程与知识工程教育部重点实验室 北京 100872
  • 收稿日期:2022-06-28 修回日期:2022-10-06 出版日期:2023-01-20 发布日期:2023-02-09
  • 通讯作者: 杨孟辉,教授,博士,博士生导师,通信作者,E-mail:yangmenghui@ruc.edu.cn。
  • 作者简介:任超,博士研究生;杨冠灿,副教授,博士;霍朝光,副教授,博士;卢小宾,信息分析研究中心主任,教授,博士,博士生导师。
  • 基金资助:
    本文系国家社会科学基金重点项目“新时期产业技术情报分析方法体系研究”(项目编号:21ATQ008)研究成果之一。

A Research on the Recommendation of Scientific Evidence in Evidence-Based Policy Based on Knowledge Graph: An Example of the Policies Fighting on COVID-19

Ren Chao1, Yang Menghui1,2, Yang Guancan1, Huo Chaoguang1, Lu Xiaobin1   

  1. 1 School of Information Resource Management, Renmin University of China, Beijing 100872;
    2 Key Laboratory of data Engineering and Knowledge Engineering, Renmin University of China, Beijing 100872
  • Received:2022-06-28 Revised:2022-10-06 Online:2023-01-20 Published:2023-02-09

摘要: [目的/意义] 为了向政策制定提供科学证据推荐,提出一种基于知识图谱的循证政策中科学证据推荐研究框架,并以新冠肺炎疫情防控政策为例进行科学证据推荐研究。[方法/过程] 提出一种有效的循证政策中科学证据推荐研究框架,该框架主要包括循证政策与科学证据知识图谱构建、循证政策中的科学证据推荐两部分;同时,以新冠肺炎疫情防控政策为例,构建包含12 872 567个实体、47种关系和61 548 684个三元组的循证政策与科学证据知识图谱,并进一步利用知识图谱的推理能力为新冠肺炎疫情防控政策的制定提供科学证据推荐。[结果/结论] 研究结果表明该方法具有良好的推荐准确率,在TransE模型中的HITS@10为0.182 588,能够有效解决新冠肺炎疫情防控政策制定中的科学证据推荐问题,也为新冠大流行期间的循证政策制定提供了一种新的方案。

关键词: 新冠疫情, 循证政策, 知识图谱, 科学证据推荐

Abstract: [Purpose/Significance] In order to provide scientific evidence recommendation to policy making, a knowledge graph-based research framework for scientific evidence recommendation in evidence-based policy is proposed, and a scientific evidence recommendation study is conducted using the policies fighting on COVID-19 as an example.[Method/Process] This paper proposed an effective research framework for scientific evidence recommendation in evidence-based policies, which mainly included two parts:evidence-based policy and scientific evidence knowledge graph construction, and scientific evidence recommendation in evidence-based policies. Meanwhile, taking the policies fighting on COVID-19 as an example, an evidence-based policy and scientific evidence knowledge graph containing 12,872,567 entities, 47 relationships and 61,548,684 triads was constructed. And it further used the inference ability of knowledge graph to provide scientific evidence recommendations for the policy formulation on fighting COVID-19.[Result/Conclusion] The result shows that the method has good recommendation accuracy, especially the HITS@10 of 0.182 588 in the TransE model, can effectively solve the problem of scientific evidence recommendation in the formulation of policies on fighting COVID-19, and provide a new scheme for evidence-based policy-making during the COVID-19 pandemic.

Key words: COVID-19 pandemic, evidence-based policy, knowledge graph, scientific evidence recommendation

中图分类号: