METHOD AND SYSTEM FOR DETERMINING SOFTWARE BELONGING BY ITS SOURCE CODE Russian patent published in 2020 - IPC G06F8/74 G06F8/75 G06F21/56 G06F21/53 

FIELD: computer equipment.

SUBSTANCE: method of determining membership of software (SW) to a certain family of programs based on its source code, in which a file containing the source code SW is obtained; determining the platform and programming language of the obtained source code; retrieving and storing code of functions present in the obtained file; deleting from stored code functions, which are library; selecting in each function a command; selecting "action-argument" pair in each command; converting each "action-argument" pair into a number; storing, separately for each selected function, the obtained sequence of numbers; accumulating a predetermined number of results of analyzing the source code and detecting therein repeated number sequences (patterns); for each detected pattern, calculating a parameter characterizing its frequency; based on the calculated set of parameters, the classifier is trained to determine the SW membership by the "action, argument" pairs sequence; trained classifier is used for subsequent determination of membership of SW to a certain family of programs.

EFFECT: technical result consists in automatic identification of software (SW) according to its source code.

23 cl, 11 dwg