Research On The Evolution Path Of Technology Topic Based On SAO Structure

Technological innovation is the driving force of the development of science and technology,and the progress of science and technology will promote economic growth.Technological path is the basis of evolutionary analysis.Technological evolution analysis can help researchers sort out the context of technological development,identify the blank points of technology and discover emerging technologies by retrospecting the development history of technology,so as to realize technological innovation.At present,the identification of technology evolution path mainly takes into account the connectivity of network,the fluidity of knowledge,the importance of chain edge(node)in network,and the similarity of text content.However,few studies have taken into account these four characteristics.In view of this deficiency,this study aims to improve the identification method of technology evolution path,enrich the content of technology path,and lay a solid foundation for the analysis of technology evolution.First of all,through literature research,the relevant theories and methods such as technology evolution path and technology topic evolution analysis are sorted out,and the shortcomings of technology path identification are emphatically analyzed.This study thinks that patent citation network's initial weights should not be the same,the text similarity with the chain on the network after weighting assigned to the importance of the network chain,as the initial weight,reuse times after calendar calculation method for the assignment of citation network for global traversal,identify the weight of the sum of the maximum path as a technical evolution path;Secondly,the improved method is compared with the traditional technology evolution path identification method to prove the effectiveness and rationality of the improved method from the practical point of view by investigating the disciplines and technical fields supported by the Ministry of Science and Technology at present.Finally,the evolution path of graphene sensor technology is analyzed,and the transition of the technology itself and the application field of graphene sensor technology from its emergence to now is summarized to provide reference for scholars in relevant fields.The specific methods are as follows:using text mining method to extract SAO semantic structure(Subject-Action-Object structure)of patent text,using neural network model Word2vec to train text vector and calculate text similarity;The SPC algorithm in the main path analysis is used to traverse the citation network,and the number of chain edges appearing in the connected path of the network is taken as the importance of the chain edges in the network.The text similarity and the importance of the chain edge in the network are weighted,and the parameters are adjusted several times by the experimental analysis method.Finally,the parameters are determined by the expert consultation method.Literature review method and comparative analysis method were used to prove the effectiveness of the improved method,and the evolution of graphene sensor technology from its emergence to now was summarized.This study found(1)the path node patents identified by the improved method almost covered the node patents of the traditional method from the perspective of path node coverage.(2)From the perspective of the integrity of technology paths,the identified technology paths by the improved method are superior to the traditional method in terms of both quantity and content richness;The time range of the path is roughly the same as the traditional method.(3)From the technical information revealed by the individual path nodes,the application field of graphene sensor technology experienced a great migration in 2004,and the key technical nodes with the transition occurred were accurately identified by the improved method,while the traditional method failed to identify them.Therefore,the technology evolution path identification method proposed in this study has a good performance in both theoretical analysis and practical results,and can be used to identify the technology path and evolutionary analysis,laying a foundation for the discovery of technology gaps and the identification of emerging technologies and other related studies.