bharat

cloaking
search engine gets a page with javascript redirecting to real page
lotka's law pr(n papers) \propto 1/n^2.
connected component sizes on the web have zipf distributed sizes
specific domains like "french pages"
"use in data structures"---?
davison classified links as nepotistic or not using C4.5 98.7% accuracy
highest similarity for same domain links ... how much of this is due to
terms that are not stop-words but result of templates, styles etc.?
50% of anchor terms appear in target page
mcbryan 1994 www worm index anchor text
* how to integrate anchor text with body text, how to weight?
surfing forward behavior studies by marchiori
bharat claims more useful for on-domain links
divide by outdegree ... not in HITS how about pHITS?
>> xerox parc information scent modeling user path prediction
bharat visualization of hits algo ... spiffy (ask for diagrams)
bharat says "precision at 10" is the most effective measure
adversarial ranking

Hi, could you please send me the softcopy of your tutorial at SIGIR 2001?


flippo menczer paper on evaluating focused crawls
seems to discriminate among links