We can be anyone online; this anonymity is good in some cases but will also be misused in many other cases. In online chats, we know that sexual predators assume a different identity when trying to get in contact with children, and they often change their age (e.g., 14 instead of 41) and gender. To prevent a large portion of cyber grooming, we can develop a framework that determines the age group (adult or child) and gender (male or female) of a user in an online chat.

In this project, we will use soft biometric voice and keystroke dynamics (typing rhythm) to determine the age group and gender of a person so that in case these do not match a user's claimed identity, we can warn the chat partner. This warning should also come as early as possible to avoid any bonding between the sexual predator and his/her intended victim.