Password Guessing Attack Based on Probabilistic Context Free Algorithm

Password is the primary way of identity authentication at present. The password security is closely related to more than 4 billion netizen all over the world. Password contains lots of semantic information, so how to extract the semantic of password and apply it to password guessing algorithm can further uncover the behavior preference of users in creating passwords, and improve the cracking rate of guessing attacks. We analyze the background and present situation of password security research, and determine the general steps and basic framework of password guessing algorithm based on natural language processing technology. We introduce the relevant preparatory knowledge and make statistical analysis on many password data sets. The popular password, password grammar, password pattern, character composition, length distribution, character distribution and semantic information of password data set are analyzed. We propose a password guessing algorithm based on probabilistic context free algorithm. The actual leaked password data set is selected for training and testing, and several groups of password guessing contrast experiments are set up. The results prove the effectiveness of proposed algorithm.