Spamming and TrustRank Theories

Mon, Aug 28, 2006

You know you’ve found gold when you see a paper refer to PageRank as “the principal eigenvector of the stochastic matrix of the Web.”

This course outline from the Data Mining class of Anand Rajaraman and Jeffrey Ullman in Stanford has some interesting sections for SEOs, particularly:

Web Crawling
PageRank and Hubs/Authorities
Topic-Specific PageRank
Web Spam

I particularly like the slides on boosting and hiding techniques, spam detection techniques and the in-depth look at TrustRank all under the Web Spam class. I’ve always been fascinated with the the Web version of trust: what constitutes trust, how is it given and how is it diluted. Rajaraman and Ullman give some clues as to how Google maybe managing TrustRank. The Google/Stanford connection only adds to the paper’s credibility I think.

Hat tip to Mike Barker from the the Web Analytics Forum.