|
#1
|
|||
|
|||
|
Seem Google will chase SE spammers with new remote workers
After all, seem Google will
http://inlogicalbearer.blogspot.com/2004/11/google-will-seek-se-spammer.html will add humans to algo, to track down spammers. Moderator note: Direct link to the job listing, which is also on craigslist: http://www.google.com/jobs/eng/test.html#qr Last edited by Marcia : 11-25-2004 at 10:19 PM. Reason: Posted direct link. |
|
#2
|
||||
|
||||
|
It can mean that or it can mean that they are reviewing the SERPs from a more general level.
|
|
#3
|
||||
|
||||
|
Quote:
|
|
#4
|
|||
|
|||
|
It is unfortunate they are not going for remote workers outside US and Canada, It they hire outside US lot will be volunteering to work, Some of the worst SE spammers are outside US AFAIK,
A spammer knows an other spammer better IMHO , |
|
#5
|
|||
|
|||
|
Quote:
The limitation (us and canadian) are probably a legal matter about remuneration laws. |
|
#6
|
|||
|
|||
|
Quote:
And definetely we Indians dont have any issue with American English, There are already many call centers and other BPO services active in India, |
|
#7
|
|||
|
|||
|
Quote:
![]() |
|
#8
|
|||
|
|||
|
IMHO it's a fact that Google have algo to track down "competitive words" in english to check keywords abuse and doorway page with keywords and link stuffing.
Make a search with "Laptop" and you'll not find easily doorway page. But now check this same query but in french with "ordinateur portable" : http://www.google.com/search?q=ordin...table&oe=UTF-8 and look the code source of the biggest company in this listing ! The same thing appear with "buy mobile phone" in french with "achat telephone portable" http://www.google.com/search?&q=acha...phone+portable In this one, you get in bonus the name of the SEO company who get recently pull out of Google for is primary domain name at the top of the code ;-) Last edited by inlogicalbearer : 11-26-2004 at 09:19 AM. |
|
#9
|
||||
|
||||
|
::cough:: cloaking humans eyes don't see what the engines do .... hmmm wild thought are they going to allow these new spammer assassins access to the Google VPN and UI interfaces.... Please post user names and passwords
![]() DaveN |
|
#10
|
||||
|
||||
|
Quote:
Quote:
Most of what's irrelevant isn't necessarily "spam" at all; I'd say the majority isn't but it still pulls down the usefulness of search for users. There's plenty of opportunity for reporting "spam" anyway, without them having to contract and pay people, between people using the the spam report and/or outing their competitors in forum posts where it's allowed - or even encouraged. Last edited by Marcia : 11-26-2004 at 10:38 AM. |
|
#11
|
|||
|
|||
|
I believe there's an interview in Mike Grehan's book, that effectively states that major search engines can be expected to have some kind of human editorial input on the bigger search terms anyway. Forget who it was he was interviewing...
|
|
#12
|
|||
|
|||
|
Quote:
Mike Grehan has a ton of good interviews in his book.
__________________
The SEO Book |
|
#13
|
|||
|
|||
|
Google does have human intervention into search results, in the sense that human will decide if particular sites ought to be banned, then those sites might be dropped. So when Google talks of "no human touching" of search results, in some sense, that's incorrect.
But Google is correct in the degree that they don't hand organize some results, make editorial picks to ensure that some sites come up at the top. MSN used to do exactly this, handpick top sites to come up for things like Britney Spears, for example. And it was actually a strength they've lost as they've gone down the all algorithm result. At Yahoo, there's evidence that for some popular queries, they are also hand-picking some results -- not just pushing buttons, but actually picking certain sites to get a boost. The version of my Going Beyond FTC Paid Inclusion Disclosure Guidelines article Going Beyond FTC Paid Inclusion Disclosure Guidelines for SEW members back in June looked at this in detail. In short, for some queries, some sites that came up had a special code that seemed to indicate they may be hardcoded to rank tops for that query. I've highlighted that below: http://rds.yahoo.com/S=2766679/K=car...=1/SS=2044565/H=1/IPC=gb/SHE=0/SIG=10pe5sku9/*-http%3A//www.cars.com/That's the redirect code for Cars.com, listed tops at Yahoo in a query on cars. H=1 seems to indicate the site was hardcoded to show up in response to this query. In contrast, H=0 means no hardcoding appears to be involved. As for this hiring, I'm with others that posted. This isn't some push to build human spam hunters. They've already got those. It looks like they want to increase the quality of results, and perhaps in particular for cases where those results may be in non-English languages. Spam elimination is part of that, but not the entire part. |
|
#14
|
||||
|
||||
|
Yahoo appears to have three levels.
H=1 Top Level Hard Coded H=2 Secondary Hard Coded H=0 Free for all No Hard Coding By controlling the top slots with hard coding, It forces the spam sites onto page 2. This has significantly driven up the PPC bids for certains terms that were normally controlled by spam. |
|
#15
|
|||
|
|||
|
Been watching the hard coding for a while now. A site seems to be able to break into a hard coded serp, but it can't maintain the ranking for more than a week or so before dropping to second page.
|
|
#16
|
|||
|
|||
|
Maybe H stands for Handreviewed/Human and the score is a measure of quality and boost factor.
|
|
#17
|
|||
|
|||
|
That is very interesting, But I am a bit confused whether it is actually human review on an automated algorithm, I checked a popular adult SERP, and the top 20 sites are given H=1 and the results from 21 is given h=0,
Most of the 20 sites are not relevant to the query and are yahoo directory sites, The fact is H=0 is the one that seem to be one relevant to the query according to the automated algorithm, But yahoo purposly places the H=1 and H=2 sites on top, That is weird, The site in 21st position is No.1 in all regional yahoo as well as altavista and alltheweb.com but it is not No.1 in yahoo.com, So yahoo seem to be more worried about yahoo.com only, ![]() |
|
#18
|
||||
|
||||
|
This is very widespread. I have checked a sampling of over 100 different terms in different industries and about 25% have the h tags for the top rankings. In general the more traffic, the more likely Yahoo touched it. Yahoo has touched some serps for pills, gambling, brand names, jewelry and other terms. But they haven't completly cleaned up any one industry. In each industry they left some serps untouched. There seems to be no pattern. Most of the terms are money terms, some are not. Sometimes they boost spam sites with redirects and some serps they only boost .edu and .gov sites. There is no connection with PPC, because they have hard coded some brand names that have no ppc. They have also touched some serps for very low levels of traffic.
It looks like Yahoo is getting an obsessive compulsive disorder over their serps. |
|
#19
|
|||
|
|||
|
Quote:
Quote:
Here's why : A 100 results query on car http://search.yahoo.com/search?p=cars&ei=UTF-8&n=100 No 48 - Volvo Cars (www.volvocars.com - 14k) : .../H=2/ A 20 results query (page 3 to get the same URL) http://search.yahoo.com/search?p=car...pstart=11&b=41 No 48 - Volvo Cars (www.volvocars.com - 14k) : .../H=0/ If it was hard/human coded, the value of H= should be the same for the same URL whatever the number of results by page, no ? |
|
#20
|
||||
|
||||
|
It appears that Yahoo is only boosting specific URLs for specific serps. They are not given general boosts. It would be very easy to not do this for searc pages with more than 20 results. Why? Because the people that change their serps to show 100 results are probably SEO people. Thus by turning this feature off it helps hide the fact of what Yahoo is doing. If I am wrong then I would like to see a H=0 rank above H=1.
or at least thats my conspiracy theory ![]() Last edited by goodroi : 11-30-2004 at 08:43 AM. |
![]() |
| Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
| Thread Tools | |
|
|