Early Identification Of Emerging Technologies In The Field Of CNC Machine Tools

Outlier patents refer to patents that have been published but have little similarity with existing patents in terms of key concepts and contents.Considering the peculiarity,some argue that outlier patents may lead to emerging technologies that are not constrained to existing technology trajectories or even be disruptive to existing solutions;however,existing research remains in examining small-scale outlier patents due to the lack of dataprocessing capacity – the recent rising of deep learning methods and big-data may help in this.Therefore,this study developed a framework for early recognition of large-scale data using deep learning model.First,based on the similarity of patent content in massive patent data.First,outlier patents are identified as candidate emerging technologies based on the similarity between patents in a large number of patent data.Second,the index of outlier patents is used as an early signal for the emergence of emerging technologies,and the number of citations of patents is used to assess the level of patent influence.Third,a DNN learning model is used to fit the correlation between early patent signals and their influence levels.Finally,the association relationship obtained by the deep learning model is extended to the future to identify potential emerging technologies in outlier patents.The key to ensuring the accuracy of recognition is the quality of the deep learning model,and the good use of the deep learning model in the field of CNC machine tool technology proves the reliability and effectiveness of the method.The framework proposed by this research is extensible,so other researchers can use this framework to explore the relationship between more early signals and impact indicators based on more data.