User Profiling for Computer System

Abstract

With the advancing technology, the use of technology for malpractices is ever increasing. It is very easy to pretend to be someone else by forging their identity or through illegal use of authentication credentials. Hence, it is always recommended to have a self-detecting system which is capable of verifying the genuineness and authenticity of the user. This project work is part of the Fraud Detection System (FDS), in which the authenticity of user is verified in real-time. Then after, this information can be used to leverage the capabilities of the system dynamically, so that the illegal user does not harm the system or have access to some confidential information. In this research work, user profiling based verification method which can verify the identity of the user in real time is proposed. In this approach, a novel method of collecting the data from the user in real time is suggested. This data is trained using recurrent neural network. With the help of this trained model, specific user's identity can be verified in real time. The verification is carried out in real-time as the user uses the computer system. Deploying a suitable mechanism to then react based on the output of the user profiling system makes the system immune to identity theft based attacks.