The PageRank of a page is calculated using a recursive formula (see infolab. stanford.edu/backrub/google.html for details) which we are not discussing here, but the underlying idea is simple. Monika Henzinger, former Google research director, explained it in an interview with the German edition of MIT’s Technology Review in April 2004 using the following analogy: Consider a doctor. The more people recommend the doctor, the better he or she is supposed to be. It is similar with ranking a Web page. The more pages that link to a page p, the higher the rank of p will be. However, the quality of a doctor also depends on the quality of the recommender. It makes a difference whether a colleague or a salesperson for the pharmaceutical industry recommends her or him. If the doctor is recommended by another doctor, that recommendation will count 100 percent; a recommendation from a nurse without comprehensive medical education will count only 60 percent, that from a patient, 20 percent, and that from the salesperson, having an interest completely disjoint from that of the doctor, will count 0 percent. The principle behind this (also found, for example, in classical scientifi c citations), is thus based on the idea of looking at the links going into a page p in order to calculate the rank of p, but to do so by recursively ranking all pages from which these incoming links emerge. The idea was fi rst explored while Google founders Sergey Brin and Larry Page worked on a project called BackRub at Stanford University. Over the years, Google has added other criteria for constructing the order in which search results are presented to the user besides PageRank. Langville and Meyer (2006) give an in-depth exposition of the mathematical and algorithmic aspects behind PageRank calculations.