Figure 1. Markov Chain Model&Figure 2. Transition matrix-Study of a Random Navigation on the Web Using Software Simulation

For a good simulation it is very important to find methods for
navigating through the web (Levene and Wheeldon, 2004). John Kemeny and Laurie Snell have
proposed the use of Markov models for web simulations (Kemeny and Snell, 1960). Cadez et al. (2000)
used Markov models for classifying the sessions into different categories for browsers. Some other
proposed techniques choose to combine different order Markov models for obtaining low state
complexity and improving accuracy, as Deshpande and Karypis (2004). Dongshan and Junyi (2002)
used for predicting the access providing good scalability and high coverage a hybrid-order tree-like
Markov model. As an alternative to the Markov model Pitkow proposed a longest subsequence model
(Pitkow and Pirolli, 1999), also for predicting the next page accessed by the user Sarukkai chose
Markov models (Sarukkai, 2000).
Transitions are simulated using the Markov Chain nodes, Google matrix and an arbitrary initial
probability distribution. Examples can be seen in Figure 1 and Figure 2.