105
User Authentication
6.3 Keystroke Dynamics
6.3.1 Basic Idea
Keystroke dynamics is the method of analyzing the way users type on a computer keyboard and classifying the users based on their regular typing rhythm. Here, users are not
only identified by their corresponding used ID and password, but their typing style is
also accounted for. In this study, people can be recognized by their typing style much like
hand writing or voice print. Recognizing typing style promises a parameter like behavioral biometric characteristics in biometric science to solve the issues in identification and
authentication technique. It facilitates nonintrusive, cost-effective, and continuous monitoring. It is a behavioral biometric characteristic which may vary during the course of a
day or between 2 days; it depends on the mental state of the person, educational level,
position of keyboard, and many more. This technique, as such, is prone to issues related to
accuracy. Thus, in order to realize this technique in practice, a higher level of security and
performance together with convenient and low-cost version is demanded with an acceptable level of accuracy. So, controlling parameters that optimize accuracy to a great extent
is necessary.
6.3.2 Science
Placement of fingers on keyboard, hand weight, hand geometry, and neuro-physiological
factors reflect on keyboard indicating a unique typing style for each individual just like
each person’s written signature.
6.3.3 Features
The features of the keystroke dynamics are Key Duration (KD) Key Latency, Up Up (UU)
Key Latency, Down Down (DD) Key Latency, Up Down (UD) Key Latency, Down Up (DU)
Key Latency, Total Time (T_Time) Key Latency, Tri-graph Time (T), and Four-graph Time (F).
Some new features can also be incorporated such as key pressure (pressure-sensitive keyboard is required), finger tips’ size (touch screen keypad is needed), finger placement on
keyboard (camera is needed), keystroke sound (microphone is needed) which were used
>8
22%
≤1
23%
>1 and ≤5
33%
>5 and ≤8
22%
FIGURE 6.5
Percentage distribution of performance (EER) recorded in literature.
User Authentication
6.3 Keystroke Dynamics
6.3.1 Basic Idea
Keystroke dynamics is the method of analyzing the way users type on a computer keyboard and classifying the users based on their regular typing rhythm. Here, users are not
only identified by their corresponding used ID and password, but their typing style is
also accounted for. In this study, people can be recognized by their typing style much like
hand writing or voice print. Recognizing typing style promises a parameter like behavioral biometric characteristics in biometric science to solve the issues in identification and
authentication technique. It facilitates nonintrusive, cost-effective, and continuous monitoring. It is a behavioral biometric characteristic which may vary during the course of a
day or between 2 days; it depends on the mental state of the person, educational level,
position of keyboard, and many more. This technique, as such, is prone to issues related to
accuracy. Thus, in order to realize this technique in practice, a higher level of security and
performance together with convenient and low-cost version is demanded with an acceptable level of accuracy. So, controlling parameters that optimize accuracy to a great extent
is necessary.
6.3.2 Science
Placement of fingers on keyboard, hand weight, hand geometry, and neuro-physiological
factors reflect on keyboard indicating a unique typing style for each individual just like
each person’s written signature.
6.3.3 Features
The features of the keystroke dynamics are Key Duration (KD) Key Latency, Up Up (UU)
Key Latency, Down Down (DD) Key Latency, Up Down (UD) Key Latency, Down Up (DU)
Key Latency, Total Time (T_Time) Key Latency, Tri-graph Time (T), and Four-graph Time (F).
Some new features can also be incorporated such as key pressure (pressure-sensitive keyboard is required), finger tips’ size (touch screen keypad is needed), finger placement on
keyboard (camera is needed), keystroke sound (microphone is needed) which were used
>8
22%
≤1
23%
>1 and ≤5
33%
>5 and ≤8
22%
FIGURE 6.5
Percentage distribution of performance (EER) recorded in literature.