Method and apparatus for normalizing non-numeric characteristics of file

The invention discloses a method and a corresponding apparatus for normalizing non-numeric characteristics of a file. The method comprises the steps of cutting at least one pair of positive instances of the non-numeric characteristics of the given file into a plurality of tokens; comparing the tokens in the at least one pair of the positive instances to obtain matched tokens; and for the matched tokens, calculating weights, matched with the given file, of the tokens and storing the tokens and the weights of the tokens in a token library.