Difficulties with the Simplified Page Rank

Although the simplified PageRank seems conceptually reasonable, it has several practical difficulties:

Dangling Links

Page et al [2, p. 6] define, ``Dangling links are simply links that point to any page with no outgoing links.''

Page observes that this arises in at least two ways:

  1. Wj points to known page Wi, but that page has no links of its own.
  2. Wj points to unknown page Wi
    (e.g., Wi has not been downloaded yet by the Google search engine)
    In this case, the link from Wj really is dangling!

Issue: Where should their weight be distributed?

Wi does not distribute its PageRank elsewhere, so any weight given to Wi will be lost in later computations.

Page et al note: "Dangling links do not affect the ranking of any other page directly."

Google's solution: Remove dangling links from the system. Thus, the only links considered within the computation of PageRank are those links that go to known pages.

created: 10 February 2007
last revised: 15 February 2007 		previous  next Valid HTML 4.01!