Slashdot Mirror
Google Propping Up Yahoo In Search Results?
c170 writes: "The position of results that point to Yahoo's pages have changed since Google and Yahoo inked their alliance." Definitely an interesting article, but not really enough statistics to prove anything ... still, definitely enough to keep your eyes peeled in the future.
That's my point. You have no right to know that. It would be nice, but you have no right to it at all.
You know what? He's correct. The source for this story had every right to analyze Google's operation in the way that he did, and he has every right to disseminate this information, and users have every right to pick another search engine. All of these rights seem perfectly reasonable, to me.
And to me. If you read my post you'll see that I suggest using another engine if you find google doesn't work for you. Nowhere do I suggest that you should not talk about their practices or methods. I maintain you have no right to demand anything from google. Anything.
Search engines perform a task that is somewhat "journalistic" in nature. In the long term, they're going to need to develop something like journalistic integrity in order to maintain the respect of most users. Giving preference to people who are paying you to do so without disclosing to users that you are doing this is probably a violation of this integrity. Not punisble by law, but certainly worthy of censure in the public eye.
Well and good, and I agree. Now please explains how this bestows any rights to the public to demand anything from google.
--
This is perhaps the most refreshing statement I have read in a comment to date. ;)
It's also entirely possible that you are correct, I still disagree, though not as vehemently as before.
Lets say that you offered Oil changes for free to anyone who wants to drive up to your house. You provide free oil, and a free oil filter. Great service. Now, after a year or two of this, you decide that you are no longer going to change the oil filters, just change the oil. (of course changing the filter, as many people know, is part of a proper oil change).
You know what, thats fine, and you have EVERY RIGHT to do that. However, the people who are getting the change have the RIGHT to know that you have changed the service that you are giving them, and it is no longer what they expect.
Perhaps, in fact yes. You can argue that safety is an issue. Cars can kill when they don't work. Search engines don't. I understand analogies aren't ideal and the point you are trying to make. I think that Google has a responsibility to provide us with this information, I don't believe we have a right to it. It's a sad fact, but not everyone lives up to their responsibilities.
It is WRONG to pretend to continue to provide a service, when you are not really providing it. This is a signifigant change in the type of service.
Again, I am in total agreement. I just do not feel we have a right to the information under the circumstances.
Noone has a "Right" to use the service. They do, however, have a right to know WHAT is being offered to them.
What is being offered is search results. Google has never made their algorithims public knowledge, and we shouldn't expect them to do so now. We (at least I) have been using google for some time now not knowing definitivly how they rank their results.
Though I share your frustration, and would love to know what's going on, I just don't feel like it's a right that has been taken away from me.
--
Correct,but if other yahoo pages linked to these, and those pages weren't indexed, their rank would be lower. Remember that google looks at how many people link to a page when determining rank.
--
Mike Mangino
Sr. Software Engineer, SubmitOrder.com
Mike Mangino
mmangino@acm.org
You mean this? Follow the Advanced Search link on the left navbar from AV's homepage.
Or, if you prefer text mode, you can get something pretty close to Raging, but with Boolean operators and date ranges.
Every day we're standing in a wind tunnel/Facing down the future coming fast - Rush
This sig intentionally left blank.
Take a look for yourself.
Bleh!
Offcourse this sort of thing is bound to happen with such an alliance.
:)
The thing to do then is offcourse to search multiple search engines simultaniously.
Which you do easiest at http://findfu.com
scary, but true... hasn't been a good day for these folks...
--
"It's tough to be bilingual when you get hit in the head."
That time was on purpose, just like this one... I'm not a chowderhead - I had Manwich for dinner, not chowder... err...
No Clevelan Steamers, thanks (I'll leave those to Marv Albert)
--
"It's tough to be bilingual when you get hit in the head."
>yeah like that comment really deserves a 2..
>who's the moron that modded this up? come clean, shit for brains.
That would be me fogetting not to use the "No Score +1 Bonus" button. Think about it.
--
"It's tough to be bilingual when you get hit in the head."
Try searching for creative apartment ethernet, you'll see nearly 3 pages of links to root pages (with a few non-root-pages sprinkled throughout).
Or search for epinions latex, it'll come up with epinion's root page, even though google's cached version of epinion's root page contains no occurances of the word "latex".
These are possible explanations:
- 1) Sites are paying google to get their home pages, and google changed their code.
Then again, if you search for basking robbins ice cream, it doesn't give you www.baskinrobbins.com anywhere, but rather gives you epinions.com/ as the third link. How the search engine came up with that, I'll never know.2) Google really really values root pages because most links are to root pages (pure speculation), so those get ranked first, and google's code sometimes ignore some of your search terms making an assumption that you messed up.
3) Similar to 2, the coders made an assumption that root pages are good, so they upped their scores artificially.
So yes, I think that google traded quality for profit.
--
On the other hand, google clearly displays information about what their algorithm does not do (see here):
- Google's complex, automated search methods preclude human interference. Unlike other search engines, Google is structured so no one can purchase a higher PageRank or commercially alter results. A Google search is an honest and objective way to find high-quality websites, easily.
So, if indeed they're taking money from other sites, then they are clearly lying. (search for creative ethernet apartment and only 10% of the links on the first three pages will have all three words in them, the rest are corporate home pages).Okay, it's mostly legal for a site to lie, but I find it exceedingly irritating when a company bases a large part of its image on being the only good one in the pack, builds a large customer base off of that, and then directly violates their image (lying is a definite way to do that).
Not illegal, but definitely hypocritical, especially for a company that asserts itself as a company that wouldn't do any funny business like be hypocritcal.
--
Yes, I deliberately misspelled it. I was also testing my hypothesis that Google might throw out a search term if it thought it could find a much better site by doing so.
--
playstation capcom software monkey (15th)
snow cedars elegance snot (1st, only hit)
gumbo pot above frogs (5th)
gerber forks babies in the ass (1st)
You'll notice that even though google lists http://epinions.com/ as a hit in all these searches, you won't find all the searched words on their root page. Google's queries search for "all words", so I don't see how it could have come up with these hits, let alone how it would have ranked the root page that high.
--
I still think Google's the best though. And this could be my imagination, so if someone else has noticed this, please post a reply.
--
Why do I have to pick one? They all serve different purposes.
If I'm looking for a very high-profile site (e.g. "Coca Cola"), or I'm looking for specific information that is likely to be out there somewhere (e.g. "Albert Schweitzer bio"), I use Google.
If I'm looking for a list of sites with a particular topic (e.g. "Everquest"), I use Yahoo.
If I'm looking for any information that is unlikely to actually be out there (e.g. dirt on my ex-girlfriends,) or very specific text that may or may not be out there (e.g. "We've got a gorilla for sale, Magilla gorilla for sale"), I use Altavista, because it seems to do the most exhaustive search.
Of course, having said all that, Google does have the nicest design by far. If Google could give me everything that I need from a search engine, I'd be happy to use it exclusively.
MSK
f I'm looking for a list of sites with a particular topic (e.g. "Everquest"), I use Yahoo.
I'd encourage you, and others looking based on topic, to actuall use the Open Directory Project first. It's passed Yahoo in size, and is usually better maintained. That, and the fact that if you see a category is rather weak, you can always sign up to update it yourself.
---
"You know your god is man-made when he hates all the same people you do."
Good for searching for geek things, or good in general?
--
The shareholder is always right.
#include "subject:"
a company that indexes data should never out-right remove data from its index unless its incorrect. I mean making that MD directory dissapear to create more hits for yahoo is counter the aims of the google engine [to get you information].
Now, maybe its financially sound for them to do it since yahoo gives them money, but its a bad trend to start since the more people who are willing to pay (and I bet there's a bunch) the less your going to be able to get from each them in the end (for example its hard to push advertising prices up unless its a special event or something). I think such a financial outlook is short termed and potentially dammaging to their reputation, which is a terribly important factor for sites like these.
-Daniel
Except that Google's page ranking algorithm (based on links to and from other relevant pages) would be useless for ranking Usenet posts. There is no Usenet equivalent of the hyperlink, only threading and article references. For that sort of data, all you need is a very fast text search.
> Claiming a right to a free service is absurd.
> Google is and remains an excellent, free,
> service. If it stops being free, or
> the quality starts to suffer -- stop using it!
Your missing my point,and thats my fault, I didn't express it fully.
Lets say that you offered Oil changes for free to anyone who wants to drive up to your house. You provide free oil, and a free oil filter.
Great service. Now, after a year or two of this, you decide that you are no longer going to change the oil filters, just change the oil. (of course changing the filter, as many people know, is part of a proper oil change).
You know what, thats fine, and you have EVERY RIGHT to do that. However, the people who are getting the change have the RIGHT to know that you have changed the service that you are giving them, and it is no longer what they expect.
When you provide a search engine, you are making the implict statment to the world that if they type in keywords or some search expression, your engine will use that expression and its own heuristics to find BEST MATCHES.
If you change your service to allo wBEST MATCHES to be "overridden" by corperate interest, then the public who uses it has a RIGHT to know that you are doing this, because you are no longer providing the service that they have come to expect from you.
It is WRONG to pretend to continue to provide a service, when you are not really providing it. This is a signifigant change in the type of service.
When you offer something to the community, be it for free or for pay, you are making a promise to provide THAT SERVICE. If you decide to discontinue offering it, thats fine. However offering a different service instead, without telling people, is trickery at best. It is just plain wrong.
Noone has a "Right" to use the service. They do, however, have a right to know WHAT is being offered to them.
--Steve
"I opened my eyes, and everything went dark again"
Stefan.
It takes a lot of brains to enjoy satire, humor and wit-
The truth shall make you fret. (Ankh-Morpork tImes motto)
Now what the fsck sort of a comment is that! Did you read the story? So its great for Google that where they were once doing a good job of finding directory sites due to their link based analysis, they are now finding only Yahoo!
Are you on drugs? Are you the Signal11 that means I'm gonna get my karma slapped?
NOTHING IN THIS IS ABOUT ADVERTISING so why are you shouting about it? This is actually bad news for Google as it appears to be demonstrable evidence that their deal with Yahoo has caused them to lose one of their competitive advantages. how long before your precious advertising limitations are gone and will that be good for Google?
Never underestimate the dark side of the Source
This isn't necessarily a conspiracy between Google and Yahoo, there is a possibility that so many people link to Yahoo that it skews the results that way.
I wonder if you read the same article I read. The one I read mentioned how the author had tested against very specific topics in the medical fields. Topics that people wouldn't link to Yahoo!, because Yahoo! doesn't really have much in the way of information about medicine. The chances that people are linking to Yahoo! on these topics is minimal, which is why the findings are so striking.
This may not necessarily show a conspiracy, but it does show that changes to Google's orignal algorithm have occurred, likely due to the new partnership between Yahoo! and Google. The article was simply pointing out that the relevancy of Google's results is not as high as it once was, due to this relationship.
So what's your point? The person you're responding to never said anything about a right to use Google. He did talk about a right to know if they're screwing around with their search result precedence to make a buck.
You know what? He's correct. The source for this story had every right to analyze Google's operation in the way that he did, and he has every right to disseminate this information, and users have every right to pick another search engine. All of these rights seem perfectly reasonable, to me.
Search engines perform a task that is somewhat "journalistic" in nature. In the long term, they're going to need to develop something like journalistic integrity in order to maintain the respect of most users. Giving preference to people who are paying you to do so without disclosing to users that you are doing this is probably a violation of this integrity. Not punisble by law, but certainly worthy of censure in the public eye.
The Inktomi-based Yahoo results can be found at HotBot, another Inktomi-based engine.
<O
( \
XGNOME vs. KDE: the game!
Will I retire or break 10K?
Actually, you can just go here and add google search buttons to your netscape personal toolbar with a click and drag.
--
As a matter of fact, I am a lawyer. But I play an actor on TV.
However, I do know that if I'm making a search in Google, I don't care to find a list of links on Yahoo as a result (else I would have used Yahoo, don't you think?).
So, to get rid of Yahoo results, simply make your search at Google Advanced Search page (http://www.google.com/advanced_search.ht ml) and put "yahoo.com in the "exclude" field.
Et voilà! No more Yahoo pages showing up in google.
-Earthling
-Earthling
"I'm sorry, I had to; the irony was just too thick."
The article raises some serious questions that some of the slashdot audience may not be aware of. Just how fair are search engine results. If you consider how much power search engines wield over the content of the internet that gets read, you'll realize just how important it is that search engines take fairness seriously. If you can bribe search engines to not list sites, you effectively have censorship. This is something that needs to be verified. If it is happening, there might be a definite need for a non-profit search engine to keep the other search engines honest.
The number you have dialed is imaginary, please rotate your phone 90 degrees and try again.
How will they make their money? This question was answered a few weeks ago here on /. The answer
One thing to remember about all this is that Google's results are also influenced by how many sites are linking to that particular website. If I were to do a search for my name on the internet, the results would first show www.fruvous.com's pages that had my name. The second would be pages at Photopoint.com. Third would be a newsgroup archive page that has a couple of my posts in it, and then my actual homepage. Why? Fruvous.com and PhotopPoint.com are far more popular to link to than www.rit.edu.
What it feels like is happening here is that the pertnership may have upped the score a little bit from sites that are lined from yahoo.com because there are more links directly from yahoo->google.
Of course, I could be wrong.
--You will rephrase your request for me to go to hell. Goto statements are not acceptable programming constructs
I read the article, and I'm familiar with the concept of how Google determins its rankings. It and www.directhit.com used fundementally different tactics then had previously been employed by the likes of Altavista, WebCrawler, and what have you. Google as I understand it uses a formula where a pages value is determined by the value of the pages linked to it. The pages linking to it's value is determined by the value of the pages that link to it tranferse X links in all directions. So for medical sites, not my forte, but any specialized subject will do, a repository of specialized knowledge which is linked to a lot of other repositories of specialized knowledge will have a high score on Google. Presumeably because each of the repositories will have had many people linking to them. Where as a lot of geocities homepages with a link to Yahoo.com will not have a high score as the individual geocities pages though great in number will not have a high scores themselves as they are not well linked to. Google does also take into account the usual stuff: meta tags, placement of keywords in the title, the body, how close together multiple keywords are etc. etc. All this goes into a "secret formula" which determines the pages ranking on a particular search. Combined with this secret formula which could well have been modified based on their recent agreement with Yahoo as suggested by this article, Google also has excellent spider technology. Now DirectHit relies on people in addition to the usual meta tags, number of occurences of the word, giving greater weight to having the word appear in the title, proximity of mulitple keywords to each other etc, in DirectHit the page that ranks the highest is the one the most people who've performed an identical search to you chose to click on. And example works best: If you perform a query on direct hit for I don't know "news for nerds" you get the benefit of everyone else who has ever performed a search on "news for nerds" if they all chose www.slashdot.org and due to the placement and occurences of "news for nerds" prominantly on the site itself slashdot could well be returned first for this querry. back to work... Muskie
(the other ones were from hotbot.com and altavista.com) Someone should spread the word about Google..
/. is a commercial entity. goto slashdot.com
Are you surprised? I'm not.
/. is a commercial entity. goto slashdot.com
In all fairness, I haven't noticed this to be the case (and I always use Google). In my experience, Google gives far and away the most useful replies to my queries, and I don't see any deal with Yahoo! changing this.
Yeah, about the time /. merges with the 'Weekly World News'. I see the headline 'END TIMES NEAR! STATUE OF NATALIE PORTMAN WEEPS HOT GRITS!'
Sigh. The David Letterman crowd wouldn't get it, but this is far funnier than anything on the list above. :)
Libertarianism is rich wolves and poor sheep playing gambler's ruin for dinner.
News for Conspirators, Theories that matter.
/. crew is posting this paranoid drivel (GPL violations, Amazon paranoia, now this) to get pageviews up because half the people fall for that shit, and another hald (me included) can't believe this is what this site has sunk to. And I am helping. Of course this is because I do work for Andover.Net Inc, but that's another story.
WTF?? "not really enough statistics to prove anything" but still on the front page, with the ever-paranoid sounding rhetorical question? where is the Geek Compound? in Area 51?
You know, what? I got me a theory: the
Sarkazmo is the assumed identity of a long-time
The problem is that they appear to be doing what many companies do. They have realised that they can make profit easier by decreasing quality.
An alliance is one thing. However compromising the quality of a service in the process means that the community of people who use it suffers.
The community thus has a RIGHT to know that this is going on, so that they can make an informed decision on whether or not to continue the use of that service.
How do you figure you have a right to anything from Google? Did you pay them? Can you point me to any written proof of such a right?
Claiming a right to a free service is absurd. Google is and remains an excellent, free, service. If it stops being free, or the quality starts to suffer -- stop using it!
I'm sure you have plenty of rights being abused that would better deserve your attention. Think about it.
--
I'll see your HTML and raise you some javascript :) Here's how to convert Netscape's "Search" button into a dialog front-end for google. Simply place this line in your ~/.netscape/preferences.js file:
i f(term)location.href='http://www.google. com/search?q='+escape(term)+'&num=100&sa=Google+Se arch'");
config("internal_url.net_search.url", "javascript:{void(term=prompt('Searchword:',''))}
I would like to see Google get into Usenet search business. I used Deja.com, but now it is a commercially bloated crap site.
With Google's extraordinary ability to sort pages, it would be an ideal move.
LinuxLover
Whenever we see results like this, we always should consider the source. Was the study done by an independent organization? Who funded the results? Was there some motive behind the study?
If you look, you will notice that the linked page is written by the people who aren't listed as high anymore. It seems fairly clear to me that they are merely bitching about not being listed as high on google, and I find it somewhat specious that google is telling their indexing software to rank Yahoo higher.
Futhermore, this could be the result of other changes designed to bring out more relevant results. For example, if they give linked-to pages a higher ability to assign relevancy (i.e. getting linked to by Yahoo gives your page a much better relevancy rating that getting linked to by my roommate) then obviously Yahoo's own pages (which are thouroughly linked by yahoo) will have a high relevancy.
There are two real possible ways to look at what happened here. Either Google put in special clauses in their code to bump up Yahoo's stats, or Yahoo adapted to the new technology to increase its scores in the search engine's data.
It would be quite simple for them to adjust their code in order to give Yahoo a leg up. All they would have to do is put in a simple if statement that would basically say "if(siteis(yahoo)) score++;" This isn't necessarily what happened, but it is quite possible.
On the other side, you must also remember that Yahoo basically is a site of lists. This is what they do, and they probably do it rather well since they've been doing it for so long -- I must admit that I don't use them all that often. When they looked at the Google data, they were probably able to update their lists to become more usefull, thus people started linking to them again, giving them more hits, they move up in the Google ratings... and this cycle continues ad infinitum.
Personally, i would say that the second case is probably the most likely one. Since they were looking for different search engine technology they were probably looking at Google and what it spat back, they then used that to stock up their lists to be better and then we have that funny little cycle I noted above. Overall, I would say there is absolutely nothing to worry about, Google isn't bumping up Yahoo's stats, Yahoo is looking at how Google works and is using the data to increase their sites value.
The article is saying, and seems to show some evidence, that since Yahoo partnered with Google that the way Google ranks things has silently changed.
Previously ranking was done based on how often a particular site was linked to, which is why google was so powerful compared to most other search engines. It was great for finding lists of pointers in specialized areas since good lists of links (i.e. a web gloss on a particular subject) would often be linked to.
Since then lists-of-links that were previously in the top 10 can't even be found in the first 500 links. Instead Yahoo comes up more often.
This isn't necessarily a conspiracy between Google and Yahoo, there is a possibility that so many people link to Yahoo that it skews the results that way. The other possibility is that the results are skewed due to the relationship between the two companies. That's what the article is about, not that Yahoo is using Google for a search engine.
For people who have actually read the article... it seems to me that what's going on here is that Google merged its database with Yahoo's, and naturally, everyone that uses Yahoo as a major resource will have links into Yahoo in their pages, and so Google's rankings have been shifted, not by 'conscious policy' but by a change in the contents of the database.
Yahoo's rise will stop when all the newly added directories have been fully spidered and statted and cross-ranked, and it'll probably fall as Google's database grows with non-Yahoo-database links being added.
Not that I have direct access to Google's database or algorithm, but, this seems more likely than a covert ranking-adjustment plan within Google.
--Parity
--Parity
'Card carrying' member of the EFF.
Yes, it happens to me too: once a couple of months ago I was trying to find the official site for a particular type of ski, and like 80 hits out of 100 where to the fscking epinions pages.
;)
I was thinking as well about epinions poisoning google, but who knows. Anyways think about this, if they made their urls fairly random, *and* had a lot of pages, *and* every one of those pages had links to the main epinions page and the main page had some dynamic link that sent you back to the other page, this might influence google's rating of the page itself.
OTOH this line of reasoning might just be a byproduct of my overworked little mind
-- the cake is a lie
If David Letterman were to do a "Top 10 Geeks Signs Hell Is Freezing Over," he'd probably need only to look at the /. headlines.
/.!"
"OMG, Compaq may be violating GPL!!! Someone ask
"Run for the hills! Google is pimping Yahoo!"
"A private company decided not to publish DeCSS -- it's the end of free speech as we know it!"
"*Red Alert* Corporations try to make money *Red Alert*"
Once Yahoo links to Google:
Yahoo users significantly increase their use of Google, and submit URLs. These URLs will be Yahoo biased, because after all, these are Yahoo users. This bias changes Google's ratings, without any other intervention.
Hell, Google was probably good largely because it was popular with geeks. Like the Net at large, it will become diluted by pr0n surfers & greed. I hope Google and Yahoo both are looking at other methods of automatic category building, since there are lots of interesting approaches to that problem.
Assembly is the reverse of disassembly.
The problem is not the making of an alliance.
The problem is that they appear to be doing what many companies do. They have realised that they can make profit easier by decreasing quality.
An alliance is one thing. However compromising the quality of a service in the process means that the community of people who use it suffers.
The community thus has a RIGHT to know that this is going on, so that they can make an informed decision on whether or not to continue the use of that service.
Such a story is wholly apropriate. Google should be contacted and asked if this was done purposfully or is coincidental, caused by some other tweak in its engine, or in yahoos pages.
--Steve
"I opened my eyes, and everything went dark again"
Herbie J.
No, no, you've got it all wrong. Though we are an alarmist crowd, the things you've described are business as usual.
The Top 10 Geek Signs Hell Was Freezing Over would go like this:
Libertarianism is rich wolves and poor sheep playing gambler's ruin for dinner.
Even better -- use this in your personal bookmark toolbar. Add a new bookmark, and then use this as the URL:
javascript:q=document.getSelection();if(!q){void(q =prompt('Enter text to search using Google.',''))};if(q)location.href='http://www.goog le.com/search?client=googlet&q='+escape( q)
Now, if you highlight some text on a web page and hit the bookmark button, it takes you to Google and executes the search. If no text is highlighted, it brings up a dialog box and asks for the keywords.
Very useful. I use it all the time.
Potato chips are a by-yourself food.
First of all Google does not rank SOLELY based on links to a page, they use a combination of the number of links to the page, the text on the page, its position, etc just like every other search engine. They also use the number of links from the page, and the text for 50 or so characters on either side of a link that links to the page. Its a wonderfully complex set of formulas that are being used to determine relevancy. While I have read the early papers on the methodology that Google is employing (from when it was a Academic project) it has obviously undergone a lot of improvements and refinements over time. They do not release the ranking criteria they are using to the general public (this is normal for Search Engine companies, who guard their criteria closely, and periodically change them without notice).
What seems most likely to me, is simply that Since Google has partnered with Yahoo, they have shared details on their ranking system or have assisted Yahoo staff in positioning the ranking of Yahoo pages in the Google database. As a result, the ranking position of Yahoo pages is on the rise simply because they have some inside information or help. That is why the pages have risen slowly over time, rather than simply popping to the top of the charts, as they might if Google had simply rewritten their formulas to make an exception when a Yahoo page is concerned.
With the work that has gone into creating Google, I do not think they would want to do any screwing around with their formulas that would result in major changes like people are suggesting here. They can help their partners rank better though.
"The first time I got drunk, I got married. The second time I bought a chimpanzee, after that I stayed sober" Arian Seid
... back in my day, we didn't have any http search tools. I had to walk twelve miles, buck naked, in the snow to find a terminal with access to an archie client, and we didn't have no graphics neither! we had to read binary code to interpret our results, if we even got any over out 110 baud connection! and sometimes the results would be in swahili, and we didn't know what to make of what we downloaded, but we were happy! not whining snot nosed little brats like yourselves, who are mad because you can't find that naked picture of Britney Spears or haven't broken 88TB of mp3s downloaded, so stop sniveling!
IBM had PL/1, with syntax worse than JOSS,
And everywhere the language went, it was a total loss...