Forum Moderators: open

Message Too Old, No Replies

Pagerank Algo! Simply 11 classes?

pagerank is only a simple logarhytmical algo

         

MOOSBerlin

5:25 pm on Oct 24, 2002 (gmt 0)

10+ Year Member



Sorry for my poor english, but could it be, that pagerank is only a simple logarhytmical algo for the classification of the over 2 billions pages/sites in the Web in to 11 classes (0 and 1-10)?
For example:
2.471.658.428 sites/pages in the web and the factor is 7.5
that means
- PR10: 8 sites (really 7.5)
- PR9: 53 sites (7.5x7.5)
- PR8: 368 sites (7.5x7.5x7.5)
- PR7: 2.573 sites (7.5x7.5x7.5x7.5)
- PR6: 18.008 sites (and so on)
- PR5: 126.053 sites
- PR4: 882.368 sites
- PR3: 6.176.573 sites
- PR2: 43.236.008 sites
- PR1: 302.652.053 sites
- PR0: 2.118.564.368 sites
That also means:
- PR9 and higher: 60 sites
- PR8 and higher: 428 sites
- PR7 and higher: 3.000 sites
- PR6 and higher: 21.008 sites
- PR5 and higher: 147.060 sites
- PR4 and higher: 1.029.428 sites
- PR3 and higher: 7.206.000 sites
- PR2 and higher: 50.442.008 sites
- PR1 and higher: 353.094.060 sites
- PR0 and higher: 2.471.658.428 sites
any ideas?

[edited by: MOOSBerlin at 6:35 pm (utc) on Oct. 24, 2002]

MOOSBerlin

9:53 pm on Oct 29, 2002 (gmt 0)

10+ Year Member



Thanks for the link, Brett, in deed very interesting!

gmoney

11:56 pm on Oct 29, 2002 (gmt 0)

10+ Year Member



Thanks for the help in msg#28 Markus.

I’ve got a couple more questions about the paper Markus referenced in msg#11 if anybody is interested in looking into the matter:

I was wondering about the constants 8*10^-14 and 3*10^-15 used in Figures 3 and 6 respectively. I am wondering what the corresponding constant would be for say a 2.5 billion page index. Any thoughts are appreciated.

Also, I was wondering how difficult it would be to get the raw PageRank data used to generate Figure 6. It appears that it is available to the public. I believe I can pay $500 in administrative expenses to get the entire WT10g database but I am really only interested in the PageRank values and I can’t justify spending $500 on my PageRank habit. I am just wondering if anybody has any connections who wouldn't mind contributing 1.69 million numbers towards webmasterworld research purposes.

MOOSBerlin

8:39 pm on Nov 3, 2002 (gmt 0)

10+ Year Member



A lot of people say, that Page Rank getting Stricter. I also think so, because google has new in the index a lot of dynamic pages (like GG said). Now the PR distributes to more pages, and some of this new pages also have get high PR. Therefore other pages lost PR!
This 33 message thread spans 2 pages: 33