A Method of Identifying User Identity Based on Username Features

Users typically register to use multiple web applications; there are many duplicate user identities in the Internet. The identification and integration of duplicate user identities are of great importance in the commercial field and network security field. As a result of privacy considerations, the personal information provided on the internet is usually incomplete or partially false. Considering that the username can reflect the user's personality or habit, is easy to obtain and does not involve the privacy problem, this article proposes a method to determine the identity of the user only by the username. First, the problem of user identity is described formally, and then the username features are divided into two categories: intuitive feature and contrastive feature, and the username probability distribution is analysed quantitatively. This article proposes a method of identity whether a given username belongs to the same user. Finally, a method is proposed to retrieve other usernames that may belong to the user in the username candidate set when a single user name is given.