There is some technique called "normalized compression distance" that does sort of it. It uses compression to compare how similar some data is to some another.

There is some technique called "normalized compression distance" that does sort of it. It uses compression to compare how similar some data is to some another. For a similar problem, you can work like it was answered here: http://reverseengineering.stackexchange.com/questions/2897/tool-or-data-for-analysis-of-binary-code-to-detect-cpu-architecture/2900#2900 http://reverseengineering.stackexchange.com/questions/2897/t...