基于主题模型的技术预见定量方法综述

Topic Modeling Based Literature Review of Quantitative Technology Foresight Methods

摘要:

基于文献计量、主题模型、社会网络分析等方法构建了一种文献综述分析框架,针对技术预见中的定量方法做了系统、客观的综述性分析,尝试识别定量预见方法与技术预见研究问题及其演进趋势。研究发现定量方法中基于现有路径预测未来发展的探索性方法仍然是主流,多局限于路径依赖的渐进式创新预见。同时,也逐渐出现了基于数据挖掘的组合型技术预见方法,能更好的识别针对未来不确定性的颠覆性技术及其路径跃迁,这也成为技术预见的未来发展方向。

Abstract:

This study uses bibliometrics, topic modeling, and social network analysis to conduct a comprehensive literature review. The aim is to provide a systematic and objective analysis of the quantitative methods in the technology foresight; it also tries to identify the quantitative foresight methods and the research questions on technology foresight and their evolutionary trends. It is found that the exploratory method is still the dominant approach for predicting the future development, which is limited to the path-dependent progressive innovation. Recently, studies use combined methods based on data mining that can identify the future uncertainty oriented disruptive technology and its path transition, which has become an emerging approach for technology foresight.