Scale-free Dynamics of Human Behavior in Personal Computer Operations

We show the dynamics of scale-free nature of human daily operations on a personal computer i.e. keystrokes and mouse clicks or wheels. A time-interval t between successive operations obeys a power-law type probability distribution p(t)proportional to t(-alpha) in a range from 10(0) to 10(4) seconds. The scaling exponent alpha is empirically shown to vary depending on variation of daily worklad n: the number of daily operations, in the form of alpha proportional to log n. With the aid of a modified fractal renewal process, we also show that the dependency of alpha on n is naturally explained if the scale-free nature of human daily operations holds for any n.