Abstract: This paper analyzes the latest research results of input and output of smart cities at home and abroad, finds that most of this field is post evaluation and lack of pre evaluation. This paper puts forward a new input-output pre evaluation analysis model of smart city, selects the input indicators by using expert investigation method, expert scoring method and principal component analysis method, obtains five core input indicators, and selects six output indicators with reference to national standards. A total of 28 cities in the first, second and third tier of China are randomly selected for model canonical correlation analysis, and finally the pre evaluation impact factor data before input and output are obtained. The results can be used as a reference tool for the pre evaluation of the construction effectiveness of new smart cities.

Key words:  smart city, input output analysis, principal component analysis, canonical correlation analysis, top-level design