首页|应用视角下基于异常检测的颠覆性技术爆发机会识别

应用视角下基于异常检测的颠覆性技术爆发机会识别

扫码查看
[研究目的]及时、准确地识别关键技术在应用领域的颠覆性信号,有利于产业的提前布局,对于提升国家和企业的核心竞争力具有重要价值.[研究方法]该研究以技术应用为导向,构建了颠覆性技术爆发机会的识别模型.首先,提取论文和专利数据蕴含的创新关系和核心语义;其次,融合市场环境与公众感知的多源数据;最后,利用BERT孪生网络模型和无监督异常检测算法识别异常点.[研究结论]以2014-2023年洪水灾害领域的相关数据为研究对象进行实证研究,共检测出142个异常点.通过分析异常点和市场爆发趋势,验证了识别模型的有效性,并总结了应用视角下的四类颠覆性技术机会,为颠覆性技术在潜在应用领域的机会识别提供新的视角和参考.
Outbreak Opportunity Identification of Disruptive Technology Based on Anomaly Detection from the Perspective of Application
[Research purpose]Timely and accurate identification of disruptive signals of key technologies in various application fields is conducive to the advance layout of industries,and it holds significant value in enhancing the core competitiveness of countries and enterpri-ses.[Research method]Guided by technology application,this study proposes a recognition model of disruptive technology outbreak op-portunities.Firstly,the innovation relationship and core semantic contained in the paper and patent data are extracted.Secondly,the multi-source data of market environment and public perception are fused.Finally,the Sentence-BERT and unsupervised anomaly detection al-gorithm are utilized to identify abnormal points.[Research conclusion]An empirical study was conducted using relevant data in the field of flood hazards from 2014 to 2023,resulting in the detection of a total of 142 abnormal points.The effectiveness of the identification model was verified by analyzing abnormal points and market outbreak trends.Moreover,four types of disruptive technology opportunities from the perspective of application were summarized.This study provides a new perspective and reference for the opportunity identification of disruptive technologies in potential application fields.

disruptive technologyoutbreak opportunityopporyunity identificationoutlier detectiondeep learningidentification mod-elmarket trendmulti-source fusion

李一铭、徐绪堪

展开 >

河海大学商学院 南京 211100

河海大学统计与数据科学研究所 常州 213022

常州市工业大数据挖掘与知识管理重点实验室 常州 213022

颠覆性技术 爆发机会 机会识别 异常检测 深度学习 识别模型 市场趋势 多源融合

国家社会科学基金重大项目中国高校产学研创新基金重点项目

20&ZD1252022IT010

2024

情报杂志
陕西省科学技术信息研究所

情报杂志

CSTPCDCSSCICHSSCD北大核心
影响因子:1.502
ISSN:1002-1965
年,卷(期):2024.43(9)
  • 17