A probabilistic model to evaluate computer software piracy

Abstract

The growing occurrence of computer software piracy has led to a new area of research, i.e., the development of methods to be used to supply evidence that software was copied.

One method to argue that computer source code was copied is to examine the occurence of strings of binary code (ones and zeroes) between the alleged parent and pirate codes. Given the occurence of a lengthy identical string between codes, and that string represents a development of executable code (versus data blocks that can be argued to exist in only one fashion), a model of the probability of repetition of such a string of code occurring between so-called independently derived source codes can be formulated. The developed probabilistic results can also be approximately by a simpler formula derived herein. A computer program and example computations are presented.