Research On Identifying Convergence Technology Of Robot Based On Text Analysis And Deep Learning

With the advancement of the “Made in China 2025” manufacturing industry and the global technological industry revolution in full swing,robotics as a strategic emerging technology is undoubtedly one of the important technical areas.Taking robotics as an entry point,vigorously developing robot-related industries is an opportunity and challenge for the transformation and upgrading of China's manufacturing industry.Under this background,it is of great significance to study and analyze the emerging technology fields that may arise in the robotics field and seize the opportunities for seizing the development opportunities in the field of robotics.Emerging technology areas birth in a variety of modes.As a new decisive factor in the emergence and development of emerging technologies,technology convergence has caused widespread concern and research.However,the existing emerging technology identification research focuses on the recognition based on the external features generated by emerging technologies.There are few researches based on the emerging technology's own production patterns and its own evolutionary laws.Therefore,how to identify the emergence of emerging technologies in combination with the emerging technology generation model is the focus of this study.In view of the defects and deficiencies of the existing research,this paper first constructed a patent text information automatic classification system based on text analysis and deep learning to complete the division of the robot field;and then used data reinforcement and deep learning methods to construct automatic classification of patent quality from five dimensions of evaluation.Finally,from the point of view of knowledge flow,the research path of patent citation network was used to scientifically and objectively analyze the process of technology integration.The patented citation network visualization and analysis software CDA was developed to design a method for recognition of emerging technologies.Based on the above research,through the retrieval processing of the corresponding data,the model was built,and the experimental validation of the proposed effectiveness of the proposed fusion-emerging emerging technology among the various sub-areas of the robot was verified,and finally the emerging technology in the field of robotics was identified.