Research On Identification Method Of Emerging Technology Based On Multi-source Heterogeneous Data

Nowadays,a series of emerging technologies represented by artificial intelligence and quantum information technology is leading a new round of scientific and technological innovation.With the increasing impact of emerging technologies on human life and economic structure,the identification of emerging technologies has attracted more and more attention.Science and technology are the primary productive force.Identifying the development of emerging technologies in the early time is the key for capturing the dominance of the development.Therefore,the research on the identification of emerging technologies is significant not only from theoretical perspective,but also from social and military perspective.In view of the shortcomings existing in the research work concerning technology identification,such as heavy reliance on expert experience,limited data source and low identification efficiency,this paper conducts the research on emerging technology identification based on the rapidly growing scientific research intelligence data.The research mainly focuses on multi-source heterogeneous data.The clue of the research is that we first obtain massive related multi-source heterogeneous data,and we carry out the standardized modeling on heterogeneous information,then we extract the technology list based on community discovery method and identify emerging technologies through multiattribute decision-making methods,at last,we conduct a case study on the field of unmanned system technology.The main work and innovations of this paper are listed as follows:(1)We propose an emerging technology identification framework using multi-source heterogeneous data.Traditional emerging technology identification methods either rely heavily on expert experience or perform analysis using limited data source,which is inefficient and unreliable.Motivated by this,an emerging technology identification framework that effectively integrating multi-source heterogeneous data is introduced in our paper.The framework consists of three components,heterogeneous information network modeling,spectral-clustering-based technology list extraction method,and an emerging technology identification method that utilizing multiple attribute decision making.The empirical analysis from UAV technology validates the effectiveness of the framework.(2)We propose a method for technology list extraction based on heterogeneous information network.Community discovery is a valid method to reveal the structural properties of complex networks,and heterogeneous information networks are proven effective in fusing and analyzing multi-source heterogeneous data.We creatively introduce the community discovery method based on heterogeneous information network into the identifying emerging technologies,consequently we propose a technology list extraction method based on spectral clustering,by estimating the length of the list and completing the elements of the technology community,we can obtain the technology list and related information in the target domain.The method can assist relevant scholars to quickly find the technology list in the target field,so as to greatly improve the efficiency of emerging technology identification.(3)We construct a series of evaluation indexes that take community characteristics for emerging technologies into consideration.At present,a major obstacle to the emerging technology identification is the lack of scientific evaluation indexes.Although traditional evaluation indexes can provide a comprehensive description for emerging technologies,they are unsystematic and difficult to quantify,which greatly lower the efficiency of emerging technology identification.We build a series of evaluation indexes consists of five indicators: scalability,novelty,growth,and uncertainty,which covers all the characteristics of emerging technologies and lays a solid foundation for the following works related to the identification of emerging technologies.To sum up,this paper proposes a complete process from multi-source heterogeneous acquisition to emerging technology identification and corresponding supporting methods,and conducts a case study for unmanned system technology.The method successfully identifies emerging technologies in the field,including UAV path planning technology,UAV mapping and remote sensing technology,and unattended sensor technology.The experimental results verify the effectiveness of the process method,also validates that it can provide guidance for the development of emerging technology identification.