The disruptive technology is of great strategic significance for the leap-forward development of China's technological innovation. In order to get rid of the situation of the lack of original innovation capabilities, and to deal with the varied key core technologies, the subversive technology identification is of great significance. This paper takes the UAV technology field as an example, reviews 2812 papers and patent data collected in the Web of Science (Wos) paper database and Derwent patent database from 2005 to 2019, with the technology based on the LDA-LSTM text classification algorithm Theme, and the CiteSpace to build a co-occurrence network, to realize the theme mutation detection from the perspective of mutation weight ranking and mutation time period and mutation co-word clustering knowledge map, and to identify the brain-computer interface technology and the gesture control in the UAV interaction The technology is a disruptive technology in this field. Finally, the validity of the recognition framework is verified through patents in the related papers in the field of the drone technology in 2020.

disruptive technology identification  /  text classification  /  subject mutation detection