PageRank assumes your browsing activity is a Markov chain in the graph defined by hyperlinks in the web (assuming 1/n probability for each outgoing link in a pa

PageRank assumes your browsing activity is a Markov chain in the graph defined by hyperlinks in the web (assuming 1/n probability for each outgoing link in a page). The rank of a page is the limit #visits to the page/#total visits (which is the eigenvector of the transition matrix for eigenvalue 1)