To date there are few researches on the semantic information of passwords, which leaves a gap preventing us from fully understanding the passwords characteristic and security. We propose a new password probability model for semantic information based on Markov Chain with both generalization and accuracy, called WordMarkov, that can capture the semantic essence of password samples. Further, we evaluate our design via password guessing attacks, on six real-world datasets, and we show that WordMarkov obtains 24.29%–67.37% improvement over the state-of-the-art password probability models. Even more surprising is that WordMarkov achieves 75.35%–96.34% attack improvement on "long" passwords, indicating the importance of semantic parts in long passwords.@en

Passwords have been widely used in online authentication, and they form the front line that protects our data security and privacy. But the security of password may be easily harmed by insecure password generator. Massive reports state that users are always keen to generate new passwords by reusing or fine-tuning old secrets. Once an old password is leaked, the users may suffer from credential tweaking attacks. We propose a password reuse model PassTrans and simulate credential tweaking attacks. We evaluate the performance in leaked password datasets, and the results show that 67.51% of accounts is breakable under 1,000 guesses, indicating our model is accurate in capturing password reuse behavior.@en