0

I am studying for an exam at university and trying to understand how I can find out if the distribution of inlinks to a page is a power-law WITHOUT knowing the actual number of inlinks.

The question that was asked last year is something like:

You are given measurements of the number of incoming links (inlinks) for all the pages in a web site. How could you detect whether the inlink distribution of web pages is a power law?

The answer given by the professor looks something like
professor's answer

I know that a method to detect a power-law is to plot the data on a logarithmic scale (such a plot is called a log–log plot) and then to check whether it is a straight line using a linear regression tool. However I really don't understand where do the numbers shown in the picture come from?

NOTE: sorry if this is not the right place to ask this question

GNUSupporter 8964民主女神 地下教會
  • 17,627
Ninjabin
  • 1

1 Answers1

0

I don't think I'm knowledgeable enough to explain it here but this chapter does a good job at explaining the mathematics principals behind your question (page 546 - test for power law). Also read up on the section regarding Preferential Attachment as it follows the same intuition as degree distribution and has been asked in previous exams. I hope it helps. Good luck with your Birkbeck exam tomorrow.

ocanica
  • 1