基于多维知识元的科学—技术关联主题识别及发展态势测度研究OA

[Purpose/Significance]Integrating scientific and technological knowledge elements within a unified analyti-cal framework for systematic consolidation and dynamic measurement proves crucial for understanding innovation develop-ment trends,optimizing resource allocation,and enhancing national innovation system efficiency.[Method/Process]This study adopts a holistic perspective on collaborative science-technology innovation,treating research papers and patents as carriers of scientific and technological knowledge elements respectively.Science-technology linkage topics serve as analytical units for knowledge element reconstruction and fusion between science and technology.The research first utilizes topic mining approaches to identify scientific and technological topics,then introduces knowledge element theory to extract key elements within topics and represent them through semantic,structural,and temporal dimensions.Subsequently,the study identifies science-technology linkage topics based on semantic associations and network structural features of knowledge elements.Considering lifecycle evolution characteristics of linkage topics,the research constructs an indicator system focusing on innovation vitality,innovation maturity,and innovation decline intensity,employing strategic positioning maps to measure and classify developmental trajectories of identified linkage topics.[Result/Conclusion]Through empirical analysis in the artificial intelligence field,the study identifies 206 pairs of science-technology linkage topics and categorizes them into 8 types including frontier innovation and emerging potential types.The empirical results confirm that the proposed multidi-mensional knowledge element-based method for science-technology linkage topic identification and development trend measurement possesses strong practicality and domain extensibility,providing systematic intelligence support for academia and industry in frontier layout planning and innovation resource allocation.