An Enhanced Non-Static Biometric Keystroke Dynamics Model for Strengthening the Information Content Security
Saleh M. Abu-Soud

Abstract
As the Internet becomes an integral part of our lives, global access to information and resources is incredibly increasing. This in turn has increased the chances of malicious attacks. In this paper, a new biometric keystroke dynamics model is presented. In our approach, the user’s typing rhythm is studied to extract the unique characteristics it has. In this model, three metrics, namely; Key press duration, Latencies between keystrokes and Key even order are combined into a new metric called DLO, which has proved to be a strong metric that showed better results compared with other models. In this research, we also added a new metric to the model, namely; the password overall typing speed which, when combined with DLO forms a new metric called DLOS. DLOS improves the results significantly. In addition, the way of calculating some measurements in the model has been improved.

Full Text: PDF