Method And Case Study On Text Mining And Main Path Analysis Based Technological Tendency Forecasting

With a new round scientific and technological revolution and industrial revolution under way,the world's nations have committed to developing their national innovation systems.Thus,the method for accurately identifying and understanding the technology development process has become an important research field.The traditional technological forecasting mainly depends on the expert decision-making,which often leads to subjective bias,opaqueness and other issues.With the advent of big data age,more and more researchers have applied quantitative methods to technological forecasting.This study improves the quantitative methods of technology topic identification and trend analysis,and attempts to build a quantitative technology forecasting framework.The achievements of this thesis are as follows.Firstly,traditional technology topic identification often uses one type of data that is one-sided,so the thesis introduced the knowledge flow law to improve it.Secondly,the classical main path analysis method based on citation network neglects text semantic information,extracts single-level path,and selects nodes excessively,text classification and technology evaluation are needed to improve it.Thirdly,from the perspective of knowledge flows,this study constructs a 5-step quantitative technology forecasting framework using literature and patent data by combining text mining and main path analysis,including data acquisition,maturity analysis,topic identification,trend analysis,trend forecasting.Take the harmonic reducer as an example,we applied the literature and patent data during 1997-2016 to the framework to predict the technology trends,which turns out that the framework is feasible and effective.At last,we forecast the technology trends for the next five years as follows: countries will devote themselves to improve the precision,transmission efficiency,and other performance-related technologies,miniaturization and light-weight of transmission,as well as optimization of component parts.