Pre-application patent quality evaluation method and system based on clustering center characterization

The invention discloses a pre-application patent quality evaluation method and system based on clustering center characterization, and the method comprises the steps: extracting keywords based on a patent text inputted by a user, carrying out the retrieval, generating a sub-data set with similar features in patent big data, and generating the center characterization of the sub-data set through a clustering model; intercepting patent information to be predicted from a patent text input by a user, and generating text representation; calculating the similarity between the text representation and the central representation of the to-be-predicted patent information, and generating constraint information based on the similarity in combination with a patent quality index; and training a patent quality evaluation model by using the constraint information, and obtaining a multi-dimensional quality evaluation result for the patent input by the user. According to the invention, mass data comparison is solved, multi-dime