Presentation Title

Using Keystroke Dynamics to Authenticate Android Users

Format of Presentation

15-minute lecture to be presented April 1, 2017

Location

IB 1015

Start Date

1-4-2017 11:30 AM

End Date

1-4-2017 11:45 AM

Abstract

In this technological era where smartphone use continues to increase, it is essential to secure user sensitive data from intruders. Many biometric authentication systems are being invented thereby not just limiting authentication procedures to password or PIN levels. Keystroke Dynamics is an emerging biometric authentication technique which involves analyzing the typing patterns of a user, hence enhancing the security based authentication. This project was aimed at identifying and verifying a user profile on an Android device, using the concepts of Keystroke Dynamics. With the rise of mobile devices, dynamic typing patterns can better contribute to behavioral pattern recognition that could not be achieved with traditional keyboards. These days, touchscreen devices allow the addition of features ranging from pressure of the screen or finger area to the classical time based features used for keystroke dynamics. When the user starts typing the static pass code, our Android App collects, extracts and conditions the data. Once the data is transformed, the binary classification model is trained to classify the user entries as true user or imposter. Our classification model is constructed on various supervised machine learning algorithms. Since one of the main disadvantages of using Keystroke Dynamics is the low accuracy rate, Filter and Wrapper Feature selection approaches were used to increase the accuracy rate of the classification model.

Department

Computing Science

Faculty Advisor

Haytham El Miligi

This document is currently not available here.

Share

COinS
 
Apr 1st, 11:30 AM Apr 1st, 11:45 AM

Using Keystroke Dynamics to Authenticate Android Users

IB 1015

In this technological era where smartphone use continues to increase, it is essential to secure user sensitive data from intruders. Many biometric authentication systems are being invented thereby not just limiting authentication procedures to password or PIN levels. Keystroke Dynamics is an emerging biometric authentication technique which involves analyzing the typing patterns of a user, hence enhancing the security based authentication. This project was aimed at identifying and verifying a user profile on an Android device, using the concepts of Keystroke Dynamics. With the rise of mobile devices, dynamic typing patterns can better contribute to behavioral pattern recognition that could not be achieved with traditional keyboards. These days, touchscreen devices allow the addition of features ranging from pressure of the screen or finger area to the classical time based features used for keystroke dynamics. When the user starts typing the static pass code, our Android App collects, extracts and conditions the data. Once the data is transformed, the binary classification model is trained to classify the user entries as true user or imposter. Our classification model is constructed on various supervised machine learning algorithms. Since one of the main disadvantages of using Keystroke Dynamics is the low accuracy rate, Filter and Wrapper Feature selection approaches were used to increase the accuracy rate of the classification model.