Keystroke Authentication Methodology Accomplish Machine Learning Model Trained Using Timing Vector.

Due to the rise of web services, user authentication is one of the prevalent areas of research. Password typing is one of the simplest techniques to achieve user authentication. However, password typing is not able to provide security. Security is achieved by using the combination of password typing and keystroke dynamics and this combination further called as keystroke Authentication. Keystroke dynamics differentiate between attacker and authenticated user by examining the typing speed. It could be a replacement of the two step verification such as biometric validation. In this paper, we propose a machine learning model which will be trained using the timing vector which was created at the time of registration of the user. This vector is the heart of our approach as it is the thing using which model will differentiate between attacker and authenticated user at the time of login. We have used various technologies like HTML/CSS, JavaScript for developing the front end of the website and backend of the website is handled by the flask based framework. The application has been deployed in a cloud based platform. For this we have used Heroku, which is an open source cloud service and it has been used as platform as service (PaaS). [ABSTRACT FROM AUTHOR]