The presented invention relates to a system and method of probabilistic passwords cracking. We describe a novel password cracking system that generates password structures in highest probability order. Our program, called UnLock, automatically creates a probabilistic context-free grammar (CFG) based upon a training set of previously disclosed passwords. This grammar then allows us to generate word-mangling rules, and from them, password guesses to be used in password cracking attacks.
The effectiveness of our approach was demonstrated by testing our tools and techniques on real password sets, several of which contained over 100,000 passwords. In one series of experiments, training on a set of disclosed passwords, our approach was able to crack 36% to 93% more passwords than John the Ripper, a publicly available standard password cracking program. Finally, the method has been optimized for speed and parallelization, because both are important requirements for any password cracking software. Therefore, the novel method significantly reduces the memory requirements of longer password cracking sessions.
