1. Advertising
    y u no do it?

    Advertising (learn more)

    Advertise virtually anything here, with CPM banner ads, CPM email ads and CPC contextual links. You can target relevant areas of the site and show ads based on geographical location of the user if you wish.

    Starts at just $1 per CPM or $0.10 per CPC.

TrustRank Q & A (part 1/2)

Discussion in 'Google' started by Michael, Jun 16, 2005.

  1. #1

    Q1. What is the first reference to TrustRank?

    The first reference seems to have been in a Stanford KSL Technical Report Open Rating Systems by R.V.Guha in 2003. At the time Guha was a Research Staff Member with IBM Research but in May 2005 he became a Member of the Technical Staff at Google.

    Q2. Was there a subsequent reference?

    Yes. In March 2004 another Technical Report from Stanford University titled Combating Web Spam with TrustRank was published on the web. This paper was later published again in a slightly different form in August 2004 in the Proceedings of the 30th International Conference on Very Large Data Bases (VLDB).

    Q3. Same author?

    No. This paper was authored by Zoltan Gyongyi, Hector Garcia-Molina and Jan Pedersen. Gyongyi was a third-year Ph.D. student in the Computer Science Department at Stanford University. Garcia-Molina is the Leonard Bosack and Sandra Lerner Professor in the Departments of Computer Science and Electrical Engineering at Stanford University. Pedersen is currently Chief Scientist for Search and Marketplace at Yahoo!

    Q4. Chief Scientist at Yahoo! I thought TrustRank was a Google thing.

    Apparently not, in the academic world anyway. There have been several subsequent papers referencing or citing TrustRank. Here are the ones I know about.

    P.Metaxas and J.DeStefano. Web Spam, Propaganda and Trust. WWW2005, May 2005.

    U.Lee, Z.Liu and J.Cho, Automatic Identification of User Goals in Web Search. WWW2005, May 2005.

    P.Chirita, W.Nejdl, R.Paiu and C.Kohlschuetter, Have we annotated enough? Research Seminar on Information Systems, Oct 2004.

    Hui Zhang, Ashish Goel, Ramesh Govindan, Kahn Mason, and Benjamin Van Roy, Making eigenvector-based reputation systems robust to collision. Proceedings WAW 2004, number 3243 in LNCS, pages 92--104.

    D.Gleich, L.Zhukov and P.Berkhin, Fast Parallel PageRank: A Linear System Approach, (Unknown date).

    Z.Gyongui and H.Garcia-Molina, Web Spam Taxonomy, (Unknown publication), Mar 2004.

    P.Chirita, J.Diederich, D.Olmedilla and W.Nejdl, MailRank: A Cooperative Approach to Email Classification, (Unknown publication and date).

    S.Chakrabarti, Web Search results’ ranking: PageRank, HITS and related work, Seminar, Nov 2004.

    W.Nejdl, How to Build Google2Google An (Incomplete) Recipe, International Semantic Web Conference, Nov 2004.

    Q5. So why does everyone think it is a Google thing?

    Because Google filed a trademark application for ‘TRUSTRANK’ in March 2005 with the United States Patent and Trademark Office. At the moment its current status remains “Newly filed application, not yet assigned to an examining attorney”.


    Part 2/2 follows....

     
    Michael, Jun 16, 2005 IP
  2. Michael

    Michael Raider

    Messages:
    677
    Likes Received:
    92
    Best Answers:
    0
    Trophy Points:
    150
    #2

    Q6. Is the filing significant?

    Probably not. Google has made 39 Trademark applications some of which like OINGO and DIRECTSPRINKS clearly have very little significance. It was probably a speculative application on the grounds that it was too similar to PageRank to allow anyone else to use it.

    Q7. Has Google commented on TrustRank?

    Not officially but Googleguy on WWW was asked “When will Google start using TrustRank”? His response was “I never try to make predictions about the future or date when the future will arrive. I especially don't try to predict the future based on an ambiguous trademark registration, domain name, or some patent filing being disclosed”.

    Q8. So what actually is TrustRank?

    Because the algorithmic identification of spam is very difficult Gyongyi et al have proposed an algorithm that has human assistance. The algorithm first selects a small seed set of pages whose “spam status” needs to be determined. A human expert then examines the seed pages and tells the algorithm if they are spam (bad pages) or not (good pages). The algorithm then identifies other pages that are likely to be good based on their connectivity with the good seed pages. TrustRank scores are calculated using a biased PageRank computation to indicate the likelihood that pages are reputable. TrustRank can then be used to either separately filter the index or in combination with PageRank and other metrics to rank search results.

    Q9. Humans classifying pages as spam. Where have I heard that before?

    Probably because Google has employed volunteer human raters and asked them to do just that. Google's Spam Recognition Guide for Raters even tells them how to do it.

    Q10. So that means Google is using these human raters to seed the TrustRank algo?

    I don’t think so because the algorithm requires only a relatively small set of seed pages and there are thousands of human raters. In my view it is more likely that Google is using the human raters to determine the effectiveness of its existing (or experimental) spam detection component in the current algorithm. However the experimental components may well include some versions of the TrustRank algorithm.

    Q11. So what’s the bottom line?

    All search engines have a major problem to solve and that is malicious spam that attempts to subvert the unbiased searching and ranking algorithms that they use. Academics and search engine laboratories are continually inventing new algorithmic ways to combat this spam and TrustRank is just one of many of the proposed methodologies. In a modified form it may be used in the future but even if it is how will you ever know? If you are a spammer then read everything you can on the subject to avoid being surprised and to develop your techniques. If you are not a spammer then you have nothing to worry about :)

    - Michael


     
    Michael, Jun 16, 2005 IP
    longcall911 likes this.
  3. digidogstudios

    digidogstudios Guest

    Messages:
    62
    Likes Received:
    6
    Best Answers:
    0
    Trophy Points:
    0
    #3
    very interesting...
     
    digidogstudios, Jun 16, 2005 IP
  4. Homer

    Homer Spirit Walker

    Messages:
    2,396
    Likes Received:
    150
    Best Answers:
    0
    Trophy Points:
    0
    #4
    Thanks Michael. I'm gonna have to bookmark this one and really give this a good read. I don't agree with your statement though. I am not a spammer yet one of my sites is banned for this reason. Here's how it happened.

    1) Insecure web form was hacked and used to broadcast spam originating in China. I was reported several times (before I knew this happened) and viola banned

    2) A web firm that I had a major dispute with decided to create false emails displaying my email header as the source.

    I am not sure which one caused the ban. But the point I'm trying to make is that you can be a good webmaster with good manners and still get nailed. Especially if you have been in the top 10 for a long time.

    TR makes perfect sense to me especially if it's confirmed by humans.


    Nice post :)


    H
     
    Homer, Jul 16, 2006 IP
  5. Homer

    Homer Spirit Walker

    Messages:
    2,396
    Likes Received:
    150
    Best Answers:
    0
    Trophy Points:
    0
    #5
    Thanks Michael. I'm gonna have to bookmark this one and really give this a good read. I don't agree with your statement though. I am not a spammer yet one of my sites is banned for this reason. Here's how it happened.

    1) Insecure web form was hacked and used to broadcast spam originating in China. I was reported several times (before I knew this happened) and viola banned

    2) A web firm that I had a major dispute with decided to create false emails displaying my email header as the source, then filed reports.

    I am not sure which one caused the ban. But the point I'm trying to make is that you can be a good webmaster with good manners and still get nailed. Especially if you have been in the top 10 for a long time.

    TR makes perfect sense to me especially if it's confirmed by humans.


    Nice post :)


    H
     
    Homer, Jul 16, 2006 IP
  6. Homer

    Homer Spirit Walker

    Messages:
    2,396
    Likes Received:
    150
    Best Answers:
    0
    Trophy Points:
    0
    #6
    OPPS: Double posted accidently...Mods can you please delete one, thanks :)
     
    Homer, Jul 16, 2006 IP
  7. Michael

    Michael Raider

    Messages:
    677
    Likes Received:
    92
    Best Answers:
    0
    Trophy Points:
    150
    #7
    Yes, I should have been clearer. When I said "If you are not a spammer then you have nothing to worry about" I was referring only to spamming the search engines not to email spam. Sorry.

    - Michael

     
    Michael, Jul 17, 2006 IP
    adamovic and Homer like this.
  8. bonez

    bonez Peon

    Messages:
    31
    Likes Received:
    0
    Best Answers:
    0
    Trophy Points:
    0
    #8
    TrustRank is a big deal in Google, more important than pagerank.
     
    bonez, Jul 17, 2006 IP
  9. Homer

    Homer Spirit Walker

    Messages:
    2,396
    Likes Received:
    150
    Best Answers:
    0
    Trophy Points:
    0
    #9
    No worries mate. Thanks for sharing your views :).

    H
     
    Homer, Jul 17, 2006 IP
  10. vaniaul

    vaniaul Peon

    Messages:
    144
    Likes Received:
    6
    Best Answers:
    0
    Trophy Points:
    0
    #10
    Michael,
    Thanks for sharing this valuable piece of information! Have bookmarked it for a thorough reading!
     
    vaniaul, Jul 26, 2006 IP
  11. netprophet

    netprophet Banned

    Messages:
    288
    Likes Received:
    1
    Best Answers:
    0
    Trophy Points:
    0
    #11
    cool stuff.........:cool:
    thanx..
     
    netprophet, Oct 9, 2006 IP
  12. Iseespam

    Iseespam Active Member

    Messages:
    44
    Likes Received:
    1
    Best Answers:
    0
    Trophy Points:
    73
    #12
    If you want to see trust rank in action search for "porno clips" in Google and look at what (url's) and sites dominate the serps.
     
    Iseespam, Oct 9, 2006 IP