融合多元方法的颠覆性技术识别研究——以类脑智能领域为例
DOI:
CSTR:
作者:
作者单位:

中国科学技术信息研究所

作者简介:

通讯作者:

中图分类号:

G350

基金项目:

:国家社科基金青年项目“基于多源知识网络的颠覆性技术分类识别方法研究”(21CTQ039)


The Identification of Disruptive Technology by Integrating Multiple Methods : Taking the Field of Brain-inspired Intelligence as an Example
Author:
Affiliation:

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    【目的/意义】当前,全球技术创新模式发生改变,对颠覆性技术的准确、快速识别需求更为迫切,难度也更高。高影响力是颠覆性技术的一个重要特征,传统基于被引频次的影响力测度具有滞后性,而基于网络位置的高影响力节点识别有助于尽早发现颠覆性技术。此外,某一技术领域中颠覆性效应往往是由多个子领域的技术突破共同催生的,而基于全领域网络的重要节点识别会忽略一些子领域。【方法/过程】本研究提出一种融合属性计量、LDA模型与网络中心性指标颠覆性技术识别方法:首先,从技术突破性和市场潜力两个维度设计评价指标,筛选出具有高颠覆性潜力的技术文献;其次,基于LDA模型将技术文献划分为不同子领域;最后,采用中心性指对分类网络中技术节点的影响力进行识别,基于特征向量中心性得到颠覆性主题子网,综合度中心性、中介中心性和接近中心进一步区分颠覆性热点主题和新兴主题。【结果/结论】基于类脑智能领域的专利实证研究表明,分类网络可以很好地识别不成熟子领域中颠覆性技术的热点主题和新兴主题,这是对全局网络识别结果的补充和完善。

    Abstract:

    [Purpose/significance] Currently the mode of technological innovation is undergoing rapid change. The need for accurate and rapid identification of disruptive technologies is more urgent and more difficult.. High influence is an important feature of disruptive technologies. The traditional influence measurement based on cited frequency has lag while the identification of high influence nodes based on network location helps to find disruptive technologies as early as possible. In addition, disruptive effects in a certain technology field are often caused by breakthroughs in multiple subfields. However, the identification of important nodes based on all-domain network will ignore some subdomains. [Method/process] A method of disruptive technology identification was proposed which integrates attribute metrology, LDA model and network centrality index. Firstly, the evaluation index was designed from two dimensions of technological breakthrough and market potential, and the technical literature with high disruptive potential was selected. Secondly, patents were divided into different subfields based on LDA model. Finally, the influence of technology nodes in the classified network is identified based on the centrality index. The disruptive subnet is obtained based on the eigenvector. The disruptive hot topics and disruptive emerging topics waere further distinguished based on a comprehensive value of the degree centrality, betweeness centrality and closeness centrality.[Result/conclusion]The empirical research on patents in the field of brain-inspired intelligence showed that classified network can well identify the basic topics, hot topics and emerging topics of disruptive technology in immature sub-fields, which is a supplement to the recognition results of the global network.

    参考文献
    相似文献
    引证文献
引用本文

邢晓昭,陈亮.融合多元方法的颠覆性技术识别研究——以类脑智能领域为例[J].,2023,(15).

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2022-12-12
  • 最后修改日期:2023-01-04
  • 录用日期:2023-02-24
  • 在线发布日期: 2024-02-27
  • 出版日期:
文章二维码

联系电话:020-37635126(一、三、五)/83568469(二、四)(查稿)、37674300/82648174(编校)、37635521/82640284(财务)、83549092(传真)

联系地址:广东省广州市先烈中路100号大院60栋3楼302室(510070) 广东省广州市越秀区东风西路207-213星河亚洲金融中心A座8楼(510033)

邮箱:kjgl83568469@126.com kjgl@chinajournal.net.cn

科技管理研究 ® 2025 版权所有
技术支持:北京勤云科技发展有限公司
关闭
关闭