Straight Out Of Left Field: Google Experimenting With Digg Style Voting On Search Results
by Duncan Riley on November 28, 2007

googlabs.jpgIf you saw this one coming, give yourself a very large prize. Google is experimenting with Digg style voting features on search results that allow users to vote up or bury search results they see.

The program, part of Google Labs, works like this:

This experiment lets you influence your search experience by adding, moving, and removing search results. When you search for the same keywords again, you’ll continue to see those changes. If you later want to revert your changes, you can undo any modifications you’ve made.

At the moment the results of the program will only be stored per user and not applied to the general search index, so that sites buried (”I don’t like”) will not appear in future results for the user, where as sites voted up will stay up. Google Labs notes that “this is an experimental feature and may be available for only a few weeks,” still, who would have thought that Google would even experiment with Digg style social voting.

Screen shot below directly from the programs site here.
googledigg.jpg

(via Googlified/ Paris Lemon)

Responses

Trackback URL

Comments

Comments Pages: [1] 2 3 » Show All

I think this would fit in great with iGoogle. There are times when I want to search for something a few times within a couple of weeks or a month and the ability to filter out certain, non-related pages would be useful.

I think Google deserves praise for this because they have been given a hard time for branching out into non-internet advertising and alternative energy. This shows that Google is committed to their core product.

 

i like this move for a lot of reasons. i find this way more interesting than the whole open social movement. google learns from me WITH my input rather than pouring through my freakin emails.

 

It’s likely that their algorithms have already implemented something similar to this through bounce rates, for example, but with non-individual results customization.

The toolbar app also has a good/bad voting system that nobody uses, although that one does not bury or dig results.

 

Sorry, but how is this either Digg-style or social? Digg has nothing to do with inventing a yes/no setting. And if my preference is only stored in my account, invisible to other users, and not impacting their engine, what’s social about it? Sounds like they’re just letting people give them feedback to the accuracy of their search results.

 

In some ways, this strikes me as a pretty logical extension of what Google already does. In this talk:

http://www.baychi.org/calendar/20061212/#2

Daniel Russell of Google made it pretty clear that they use user behavior to affect search results. Effectively, that’s using implicit voting, rather than explicit voting.

The personalization based on user interest isn’t new, either:

http://blog.searchenginewatch......628-073541

I think the main difference is the extent to which this is all in the user’s face. Given that a lot of people love Google for the magic simplicity, I’m not sure that’s a step forward.

 

Somehow I agree with Dan. This seems to be a feature so a single user can taylor his/her own search experience when doing future searches. I don’t see the social element anywhere. Yes it could be extended to make it more open, but as it stands this is not a “social” feature.

As for associating anything that has to do with voting to Digg, well, that’s another story :-)

 

I completely agree with Dan’s comment; It’s neither Digg-style nor social, but it’s a nice experiment nonetheless.

 

My money is that digg will open their APIs to allowing embedding of Digg technology around other types of data. It’s only a matter of time before you will be able to digg anything.

 

Dan
vote up or bury. Sounds like voting to me, which is what I did last time I was on Digg, or even Reddit and other Digg-style sites.

 

I would take Google over Digg any time of the day.

 

The big challenge will be for Google to prevent people from gaming the results. This opens up a lot of possibilities for someone in control of a bot net.

 

I actually had an idea for a search engine a year ago that included social elements and more, drafted up the business plan and everything, but pulling together a team can be tough. I’ve got a number of friends in the Silicon Valley, but they can be lazy and very negative if the idea doesn’t come out of their uncreative, ‘rational / logical’ and boring minds.

 

Simply a clone of the http://www.sproose.com concept.
So the large prize goes to the sproose.com guys, they saw that coming 2 years ago.

 

If using digg, how about SEO?

 

I think this must be a most useless feature offered by Google, ever!

I’d kill my self if it ever comes out into mainstream search.

 

@duncan: yes, digg and reddit have up and down, but the key is the social. when you hit the up arrow, you are promoting that story for all the other users to see. your post specifically states that google is not doing this (”will only be stored per user and not applied to the general search index”) so i fail to see how this is digg-style or social. it doesn’t sound like social voting at all, it sounds like the google equivalent of a highlighting your notes.

 

Hmm, if this got opened out, Digg-style, it looks like something Subvert and Profit (or someone similar) would jump on to try and game for SEO benefits…

 

I heard hotornot is adding a digg-like feature, but with a voting “scale” rather than just binary options. I can’t wait. But why did they need to “copy” digg?

 

Cool now Online Debris can be avoided. Hope they use decisions of influential people collectively and make them available to common people.

http://tekno-world.blogspot.com

 

I’ve always wanted Vi-like style shortcuts on Google, I already use them extensively on Google Reader and GMail, I was wondering when would they start enabling them on Google search as well.

 

This is a nice little thing from Google. Sometimes the search results contain crap or porn which is not intended. All those tricky MFA websites and cloaked websites can now be “buried” for good.

I don’t really hope to see the feature any time soon. I don’t know many people who are logged in to their iGoogle while surfing. Most are just trying to find something from the toolbar.

Lets wait and watch…

 

Its a good idea but I don’t think it can work without being gamed. Wouldn’t be to hard for the bot commanders out there to game it. I suppose it would only be an extra field on their page rank calc and thats up to well over 120 odd.

 

Our organization has been working on a similar but more advanced concept for months. Our approach works with all of the major search engines and does provide a far more sophisticated mode of operation rather that just ‘like / don’t like’ and ‘bury’.

 

I think it will just affect your own results. However if a site gets hundreds of buries they would probably flag the site and the search results page for a manual review. Does this mean for all Google users? So if a page maybe gets 100 thumbs down then Google remove it? Not sure??

Parul
http://www.bhopu.com

 

“If you saw this one coming, give yourself a very large prize.”

What do I get? A cuddly toy?

I talked about Google taking on Social Media smarts some time ago.

So for me, it’s about time Google finally ‘got’ Social Media…

 

I totally agree with Subvert and Profit , very easy to manipulate for SEO purposes.. I mean what is it with you diggers, some of you seriously think you own the internet.. there is nothing more than diggers love than talking about digg… get in the real world , google is doing this so it can improve its results… as already said above I would take Google over the sexually frustrated users of Digg any day.

 

About two years ago I predicted that this would be the future of search. It makes sense when you sit back and think about it. Google’s algorithm is supposed to work by measuring the number of links pointing at a site. This is essentially a vote. Their algorithms try to predict the votes which are a better representation of quality by building connected graphs of “quality” web sites. There are two other ways to determine votes of quality. The first is to measure the time a user stays on a page. This could be determined by 1. tracking when a user leaves Google and when a user returns, 2. by measuring page stay through Google ad server requests, or 3. by buying up large pipes and monitoring traffic. All of these methods seem rather big brotherish. They may be used but it would be best to do so out of the scrutiny of the public. The other method is to allow a person to actively vote for the sites he or she likes and/or downgrade the others. I will now give myself a large prize…(a.k.a. grande coffee before work.)

 

I hope Google clarifies if the dig/bury system is for relevancy or for ethics because if people think it’s for relevancy and start burying quality web sites that aren’t elevant for their particular search, a lot of good sites may be added to the black list.

 

http://www.irazoo.com has implemented this Social Search feature early this summer, but they’re using to to refine all of the searches for that topic based on human-powered relevancy - without relying strictly on algorithms.

 

I’m so curious about how people will go on spamming the searchresults. Personally I think black hat seo will concentrate even more on small niches so there won’t be to many people who rate the sites bad.

 

WooT! I can finally get rid of the *.about.com links! I love it! :)

Later…
Richard

 

Finally we can all remove Experts Exchange from our search results!

 

At the moment the results of the program will only be stored per user and not applied to the general search index, so that sites buried (”I don’t like”) will not appear in future results for the user, where as sites voted up will stay up.

In other words this will be applied at the personal level, and have no impact on anyone elses search results for the same search text. Therefore it is not a social implementation, but a method to control what you personally see in the results.

Its only going to work for search results that return the same sites - otherwise it would be a pointless exercise. It appears to me to be a way of getting rid of sites that you dislike from appearing. I can’t believe this would have any impact long term on the search results presented across the board, otherwise everyone would vote down the competition.

 

To bad theres no way to promote a google cache result to punish but still use the pages that show the spider a different page. A heat map in the background of a link would be the obvious choice.

 

Isn’t the best voting is simply following what users click? Obviously some clicks do not supply the best answer or information, but it should give some idea of whats useful and whats not.

How many people will go back to the search results page to vote after they already found the information they needed?

 

I believe it has to be tightly integrated into existing social networks. Ideally, search results should not be influenced by the whole internet community, but only by the preferences of the close friends group.

 

I would say this is a better fit for Mahalo. It could help the editors there in a big way.

 
 

This will enhance spamming of serp. People will vote on their own sites like digg

 

God is on the war path, I suspect the agrodated data of something like this will be around and developed for a long time to come. God has long since recognized that their algorithms are game-able and they surely could improve the results with a large flux of data gathered from the wisdom of crowds.

This might be timely :
http://startupcrunch.org/who_w.....extlinkads

 

what about searching only sites that appear on digg reddit like http://www.50matches.com does?

 

Good for them. It’s interesting to see them making major changes in the way pages are ranked.

 

This could be a great thing. To touch no the topic of SEO, this would be interesting to see what happens.

The chances of you being able to manipulate results in an non-ethical manner aren’t likely since you have to be logged in to use the voting system. I see it working to an advantage because it gives the chance for logged in users to weed out the junk and bring the goods to the top. In return, this should drive positive traffic to your website and that would effect the general populations search results (users who aren’t logged in)… Right?

Now, I haven’t gone through and read up on more details regarding the experiment, but wouldn’t it be awesome if it had an “A.I.” that naturally returned sites based off of your voting habits? If this happens; please mark my words and give me credit as needed.

HA!

-ronlg

 

Off hand I dislike this change, if feels like a pure popularity contest. On the other hand I will gleefully mark down required registration sites like “Experts Exchange” or “NYTimes” and useless sites like “BizRate”.

 

Vertical search engines, and one in particular — Wazap have had this feature for over a year, AND it’s social, AND it effects the overall index weighting of the result.

 

First there was Web History: Google stores your search terms in private
Now there is Google Labs: Voting on search results in private
What will be next: Making both social? I think so! In my point of view the integration of search and social networking could make sense…: to improve search fulfilment and to improve networking…

 

All hail the Google Bury Brigade!

 

Ohh finnaly i get to take out the stupid search results i dont like :)

 

We’re missing the point that folks, especially bloggers, are not going to hang out on Google to promote stories. I see this as Google testing the waters - maybe to purchase a Digg site in the next 3 years. sssh, don’t tell anyone.

 

Thank You Duncan. I’ll take one large prize thank you.

Perhaps Google is in fact experimenting to see the ‘how’ and ‘what’ users do, to better write their next custom search algo.

Gord Hotchkiss discussed this in August; http://ask.enquiro.com/2007/go.....ted-world/

As a search user, I would hope Google doesn’t take a tiny sample (”a few weeks”), and apply those standards to general results.

I get it that user input can’t be wrong. But it can when the user eventually doesn’t even know what other choices could have been available.

And then there’s the risk of a gamed system if it is applied to general results.
A couple thousand thumbs up for your link for $50, from a third world spam farm anyone? (Let alone the blackhat automated software programs that will be attempted).
I’m sure Google is aware of all this.

 

Comments Pages: [1] 2 3 » Show All

Leave a Reply

« Back to text comment