A statistical Markov-based password strength meter

Although multi-factor authentication is gaining popularity, password-based authentication remains the most commonly employed method for both online login and data encryption. To help users choose secure passwords, password strength meters (PSMs) are a well-known and important tool. However, many PSMs still use simple rule sets or rely on heuristic results. With the continuous development of password-cracking methods, it is difficult for such PSMs to provide accurate assessment results. In this paper, we present a new PSM based on statistical results from a huge number of leaked passwords. The proposed method has a very simple structure, requires only a small amount of storage space, and succeeds in providing reliable feedback in real time. We also confirm the influence of linguistic features on user-created passwords and, therefore, recommend evaluating passwords based on the users' language.