Linguistic Test of Chinese and American National Spirit based on Speech Recognition and Artificial Intelligence

In this paper, the subband analysis method is introduced, and a new multi-feature subband hybrid model is proposed. This paper first proposes a method to use the user-confirmed text in the process of human-computer interaction to help improve the accuracy of self-adaptive data machine labeling; the paper selects the typical American "Social Learningʺ of Western national spiritual education and my country's "People's Education Editionʺ. The content of national spiritual education in grades 3-6 in Morality and Society was tested in linguistics, and some useful enlightenments for the current national spiritual education in our country were put forward. Then an acoustic confidence method based on the confirmation model is proposed.