Using data stretching over more than 5 years for the website http://gallery.future-i.com/ this paper investigates basic properties of the popularity of pictures and the way users navigate through the website. We find that the rank frequency plot of the downloaded pictures follows a Zipf law. The download rate of individual popular pictures over time resembles that seen of infection rates of diseases. The graph created by successive downloads of pictures shows a power law that does not change over different time periods. We discuss how our findings can be modelled and how they are of importance to website performance.
pubs.doc.ic.ac.uk: built & maintained by Ashok Argent-Katwala.