A Technology Convergence Prediction Method Based on Graph Neural Networks
[Research purpose]Accurately predicting technology convergence trends helps enterprises to advance their industrial layout,discover technology opportunities,and enhance competitiveness,thus lead industry development.[Research method]This study propo-ses a technology convergence prediction method based on Graph Neural Networks.Firstly,Named Entity Recognition is used to extract technical terms from patent abstracts,and a term co-occurrence network is constructed.Then,vectors are generated for each term as node features,and the GraphS AGE algorithm is employed for link prediction.Finally,social network analysis methods are used to analyze and interpret the link prediction results,obtaining technology fusion prediction outcomes.This study conducts empirical research using the field of Natural Language Processing as an example.[Research conclusion]Empirical results indicate that the proposed method can uncover more novel and important nodes and predict technology convergence relationships associated with these crucial nodes,providing a solid ba-sis and inspiration for technology layout and R&D.
graph neural networklink predictionnatural language processingtechnology convergencesocial network analysis method
万方数据
维普
