ESR 1: Usability Performance Assessment of Mobile Touch-Screen Behavioural Biometrics

This project has assessed the quantity of discriminative behavioural biometric enrolment data – the user usually desires an efficient enrolment process with minimum data sample donation taking less time. In this research, a method to efficiently authenticate a user using mobile device touch interaction data with a minimum number of data donation samples was identified. The outcome of this analysis can inform the biometric system design of the optimum number of samples required in order to accurately verify a user using these modalities (swipe, signature and keystroke dynamics).

What was achieved and the impact:

A multi-modal solution was developed with a combination of balancing user’s needs while efficiently identifying the user. This solution provides enhanced security compared to individual modal solutions as it is difficult for an attacker to replicate the behaviour of all the modalities.

Authentication performance assessment for different combination of behavioural biometric modalities was conducted. This analysis indicated key challenges with different combinations and revealed the best combination with respect to the performance and applicability.

Published work can be seen here.