Article Preview
TopIn this paper, (Kim & Kang, 2020), Kim and all propose a new KDA method based on freely typed text for mobile devices named FACT. As a first step, they are developed a typing data collector from three different smartphone sensors in two languages (English and Korean). From the raw data, they have extracted 17 micro behavioural features related to finger length, hand size, finger size, and muscle flexibility. The evaluation results of this approach have shown perfect protection ability with Korean. This typing data set is made public without any restrictions.
The work of (Buker et al, 2019) analyses the interaction between gender and typing dynamics. The experiments show that when gender is different, people tend to type differently while interacting with live chat interfaces, such as those available in popular apps, like WhatsApp or Skype. The experimental results show that the differences, in terms of gender, are sufficiently consistent to allow automatic gender recognition with an accuracy of over 95%.
In the work of (O'Gorman, 2003) the attacker looks for the password while it is being typed. One of the best solutions for all these attacks is to combine the biometric pattern with the password. By doing this, even if the attacker looks at or knows our password, he cannot type in the same way as legitimate user types.