Catch up on stories from the past week (and beyond) at the Slashdot story archive

 



Forgot your password?
typodupeerror
×
Google

Google Personalizes Search Results Even When You're Logged Out, a DuckDuckGo Study Finds (theverge.com) 114

According to a new study conducted by Google competitor DuckDuckGo, it does not seem possible to avoid personalization when using Google search, even by logging out of your Google account and using the private browsing "incognito" mode. From a report: DuckDuckGo conducted the study in June of this year, at the height of the US midterm election season. It did so with the ostensible goal of confirming whether Google's search results exacerbate ideological bubbles by feeding you only information you've signaled you want to consume via past behavior and the data collected about you. It's not clear whether that question can be reliably answered with these findings, and it's also obvious DuckDuckGo is a biased source with something to gain by pointing out how flawed Google's approach may be. But the study's findings are nonetheless interesting because they highlight just how much variance there are in Google search results, even when controlling for factors like location.
This discussion has been archived. No new comments can be posted.

Google Personalizes Search Results Even When You're Logged Out, a DuckDuckGo Study Finds

Comments Filter:
  • by ShanghaiBill ( 739463 ) on Tuesday December 04, 2018 @11:18AM (#57747590)

    Obvious solution: When you need to buy drugs, hire an assassin, or process your bitcoin payment from the Russian FSB, just use someone else's computer. I use my cubie-mate's while he is taking a toilet break.

    Another option is to use the terminals at the public library. Just watch out for the security cameras.

    • What is a cubie-mate?
      • What is a cubie-mate?

        Someone with whom you share a cubicle when your employer is too cheap to provide each employee their own.

        • by recjhl ( 840587 )

          What is a cubicle? :-)
          I have never seen one in use in Denmark.

          • I have never seen one in use in Denmark.

            Then how do Danish CEOs pump up the value of their stock options with short term cost cutting?

  • My observation: In recent years, Google has been, more and more, poorly managed.
  • Not surprising (Score:5, Interesting)

    by cyberchondriac ( 456626 ) on Tuesday December 04, 2018 @11:21AM (#57747624) Journal

    Everyone's watching, listening, and logging all the time. It's creepy.
    The weirdest example to date: Just this Saturday, I came across a meme on the Memedroid app on my tablet, about a nerdy hoodie that looks like knight's armor. Some comments were pro, some con. I moved on to the next meme.
    An hour later, I went downstairs, on my PC, check in on Facebook.. guess what an ad for shows up. That hoodie.. that I had never seen before that meme, and certainly never searched for .. anywhere, ever.
    WTF? I have FB installed on the tablet but it wasn't actively running. That shit is spooky. Neither that meme nor in the comments for it was a link, it was just a picture of the stupid thing, and a joke. AI ?
    I cannot bring myself to believe that was pure coincidence. It's one example of many, but just the most egregious.

    • Re: (Score:2, Informative)

      by Anonymous Coward

      "I have FB installed on the tablet but it wasn't actively running." Yeah, it was.

      https://www.theguardian.com/technology/2017/jul/03/facebook-track-browsing-history-california-lawsuit

      https://thenextweb.com/google/2018/08/14/google-is-tracking-your-every-move-even-when-you-tell-it-to-stop-heres-how-to-fix-it/

    • Re:Not surprising (Score:5, Informative)

      by E-Rock ( 84950 ) on Tuesday December 04, 2018 @11:30AM (#57747694) Homepage

      Can't speak to what that one does, but ad networks pull cookies and do other fingerprinting. So Facebook wasn't open, but an ad network cookie was there, so it could get your ID and feed it into the network. Boom, ad can now connect to you all over the place.

      It's super creepy.

    • by tsa ( 15680 )

      That's nothing. One day I was just thinking about something that I not normally think about. Later that day I looked up the YouTube main page and there was a movie about what I thought about in the Recommended section.

      • by cascadingstylesheet ( 140919 ) on Tuesday December 04, 2018 @12:04PM (#57747918) Journal

        That's nothing. One day I was just thinking about something that I not normally think about. Later that day I looked up the YouTube main page and there was a movie about what I thought about in the Recommended section.

        Oh yeah? That's nothing [dilbert.com] ... YouTube shows me things even before I think about them!

        • by ffkom ( 3519199 )

          That's nothing ... YouTube shows me things even before I think about them!

          That's nothing! YouTube shows other people things way before I have ever thought about them!

    • I'll tell you what's creepy. There's a certain restaurant that I frequent in LA. I'm typically out there a couple of weeks every year doing some work. I've never given that restaurant any personal information. Ever.

      A couple of days ago I got an email from them, and they've added me to their email list. I'm still not sure how.

    • I saw no meme and in my experience, that hoodie absolutely flooded the internet last week, for some reason. So it's still possible that it's a coincidence.

    • by jred ( 111898 )
      I've had an in-person voice conversation with a friend discussing a moderately unknown religious guru. The next day I started seeing ads for their retreat...
      • I've had an in-person voice conversation with a friend discussing a moderately unknown religious guru. The next day I started seeing ads for their retreat...

        Colocation based advertising. You were near the guru location (probably close to an advertised date time and location or an appointment on someones google calaneder ) therefore googles ad network assumed you were interested in the guru.

    • by Waffle Iron ( 339739 ) on Tuesday December 04, 2018 @12:09PM (#57747952)

      It happens on this site, too.

      I look at one story, and most of the posts are trolls about APK, immanent Trump incarceration, and giant swastikas.

      Then I open an article on a completely different topic, and what do I see? The very same posts about APK, immanent Trump incarceration, and giant swastikas.

      It's downright creepy.

      They even do this down to the micro-level, randomly inserting "â(TM)" into people's posts on my browser, no matter what the topic. I assume that they're targeted promotions for this trademarked "â" product. I don't know where they got the idea that I was interested in â.

    • If an app is free, then you're the product. Free apps tend to want more access than they need, for that reason.

      I can't bring myself to believe that was pure coincidence either. Try visual search on bing for an example of how this technology works to match images with products. Also keywords.

      • There is a free version of the app, but I use the paid ad-less version (it was only like $2 or $3 and got rid of the ads).
        It could still be that the developers just pass everything along to people like Facebook and other 3rd parties, regardless.. I'd rather that, frankly, to having my screen monitored 24/7 regardless of what app I'm using just because I've installed FB on the tablet. (And I only use it on the tablet very rarely).
        Monitoring me to try and sell me crap is one level; it's a little creepy but

        • "what I'm really concerned about is, when (not if) the government decides to join the party."

          How do you think Faceboot and Big Brother Google make all their money? Selling ADS? Yeah, right......

    • not only is it not a pure coincidence, but you might not be too surprised to find out that Tenor Gif has been long acquired by google, in order to serve ads to users browsing memes, by tagging said memes full of advertising keywords.
    • by tlhIngan ( 30335 )

      I have FB installed on the tablet but it wasn't actively running.

      Yes it was. Most likely your app was a webview that displayed a website complete with Google +1 and Facebook Like buttons.

      Those buttons are your trackers.

    • I have no Gmail account. I have no Facebook account.

      1) Once, I Google-searched some work-related part using my work desktop PC, and that night, back before I blocked ads on all my machines, an ad for that part showed up on my home machine even though I didn't do the same search at home.

      2) I once looked unsuccessfully for third-party ink cartridge refills for my 20+ year old Epson inkjet printer, since Epson stopped making them. A couple of days later, I got an expensive, glossy, multi-page ad in snail
    • If you install an app, and give it access to your browsing history, it is storing all of it.

      If you install an app, and give it access to your microphone, it is listening at all times and telling somebody.

      If you install an app, and give it access to your location, it is tracking all your movements and selling them to a company that correlates it with all your other data, and then sells it to everybody.

      It is not a coincidence, it is the natural result of clicking "yes" when asked to share your browser history

      • You're not getting it. Memedroid is an app, but even if it had access to my browsing history, I didn't browse to this hoodie. There was no "Browse history" to speak of. It was simply a picture of one as part of a meme on this app. As I stated, no URLs, links, or mentions by exact product name were referenced.

  • ... yet I get personalized results. It is kind of freaky...
  • I knew that for at least a year, just from experience. I thought it was common knowledge.

  • by Solandri ( 704621 ) on Tuesday December 04, 2018 @11:58AM (#57747874)
    Except for my email (which runs on a browser in a virtual machine), I browse completely in incognito mode. I notice searches start to become biased depending on what else I've searched for or browsed in that tab. The fact that the suggested search terms (which pop up as you type in your search request) seemed to "know" what I was browsing recently was a pretty big clue what was going on. Closing the tab and running the search again in a new tab clears this up and reverts the search to its default (which sometimes means different search results compared to the old tab)
    • by ntropia ( 939502 )
      (Apologies if it might look like I'm trying to turn this place into a venue for having useful conversations)
      Isn't that a bit of an overkill? I can think of a number of alternative options:

      - running multiple profiles for different activities, i.e., Firefox/mail profile and Firefox/browsing profile

      - running different browsers for different activities, i.e., Firefox for the email, Chromium for incognito browsing

      - running different browsers/activities as different users, i.e. "sudo -u mail_only firefox ww
      • I use Firefox containers and they mostly work. The default one is for random stuff since that's where clicked links open. I created one for work and one for Facebook and I try to stick to them (at least the Facebook one). I also run ad block, but usually allow the first party site (even FB). I notice that it seems to serve only adds about tech doodads I click on from friends posts, which is much less annoying. I think they really want to serve you stuff you may actually shell out for, so you can work a b
  • From the DDG study:

    Second, search results can change by location, such as the inclusion of local news articles. We controlled for this factor by checking all links by hand for this possibility, comparing them to the city and state of the volunteer. We saw very few local links for gun control (1 organic link, 1 news infobox link) and immigration (0), though more for vaccinations (15 organic links, 4 news infobox links).

    To control for these local links, we replaced all of them with the same placeholder — localdomain.com for organic links and "Local Source" for infoboxes — in all of our analysis. This adjustment means two users whose results only differed by a different local domain in the same slot would not count as different. Interestingly, this adjustment didn't affect overall variation significantly.

    Unfortunately, that doesn't really control for location, because the targeting doesn't work the way they think it does. Google doesn't just include local news stories, but, even for (especially for?) logged out users, they apply targeting based on what your local demographics are like and the search history results of your neighbors. Live in a big city? Even if you're logged out, you'll get a different set of results than if you live in a small rural town. This is true even with a completely wiped history or brand new computer. The justification is that you probably have many similarities with people around you... if they're all searching for snow blowers because there's a storm coming, you probably are interested in one too. It's not even close to 100% accurate, but it's not inaccurate either - it's the same basis used for decades for selecting markets for television commercials, too: using a small group of consumers for whom they have highly accurate information, they extrapolate out to the larger market.

    Does this mean you're not really logged out, and Google is secretly tracking you? No, no more than you're being tracked when some broadcaster decides to show certain commercials during a sitcom as opposed to others. They're just making an educated case, and while the result looks the same - pseudo-personalized content - the process is different.

    • I would assume that part of their point is their method might not completely control for location. The point of the study is to prove (not just assert) what ways google is adapting results in non-obvious ways. So they try to control for things that are fairly obviously localized, since those are "expected" differences. I think a major point is that google could offer an "unprofiled" search mode. But not only do they not offer that, they bait-and-switch by offering something that can be easily mistaken f
  • by hey! ( 33014 ) on Tuesday December 04, 2018 @12:32PM (#57748086) Homepage Journal

    Online news sucks specifically because it is excessively tailored for you.

    One of the lost pleasures of 50 years ago is reading the paper; modern papers are ghosts of their former selves. A newspaper was a carefully curated collection of informative articles designed to appeal to a broad variety of people in a geographic area. Yes, they had ideological focuses, but narrow that focus too far and circulation would drop. Because newspapers desired the largest possible audience within a restricted geographic area, items in them had to stand up to critical scrutiny from a number of points of view.

    Since there were no smartphones, when you had a little down time you'd read a bit further into the paper until you were scraping the bottom of the barrel. I'd start with the front page, go to the science section and work my way down until I was reading the sports page. And when you finished reading you'd be just a tiny bit different than when you started, because you'd been exposed to unfamiliar issues and viewpoints.

    That feeling of having your mind expanded is what I miss. You can spend a few hours reading online news but when you're done you won't be any different than when you started. While you're reading you may be entertained, provoked, and pandered to, but in the end the algorithm isn't there to inform you. It's there to pigeonhole you so you can be bundled for sale.

    • Online news sucks specifically because it is excessively tailored for you.

      That's a blanket statement that doesn't apply to everyone and in all situations. If you had said "corporate curated newsfeeds", you'd have pretty much nailed it.

      I do mostly online news. But I do it with an RSS feed that pulls in a fairly diverse set of news websites, including local news. I've curated political, science, and tech news sections, and I've tried to do my due diligence and put a good bit of diversity in there, provided that the sites in question are actually generally factual and truthful.

      What

    • Your local newspaper is tailored to you, so is all your historic news, simply because you have a lot in common with people living around you

      the only thing that has changed is how tightly focused this is - turning off cookies stops the tight focus and they can only tailor content to you based on context ...

  • I have people I know who LIKE walking out of a restaurant and being asked to fill out a survey. They buy into the tracking / convenience BS. I don't understand it at all, it would drive me nuts. But honestly I don't see it very often at all. Here is my setup:

    1. Home PC I run Linux (Devuan), and use PaleMoon. Yes, I use google as my search, but I am not logged into my google account on my PC.
    2. Android phone, logged into my google account, and I get my gmail there. I only use that email for "official"

  • It's not clear whether that question can be reliably answered with these findings, and it's also obvious DuckDuckGo is a biased source with something to gain by pointing out how flawed Google's approach may be.

    Thanks for being open to criticism and openly acknowledging faults/bias.

  • by Anonymous Coward

    The DuckDuckGo owner goes around *constantly* bashing Google, sometimes dishonestly, in every single interview he gives. I mean *every* interview. There's no such thing as DuckDuckGo doing a "study" of Google.

  • Incognito mode just guaranteed Google won't spy on you. It doesn't mean the web site isn't sending what you click on and your IP to an advertising preferences database (Google, Amazon, facebook), and Google is smarmy because their own sites do this even though you are in incognito mode and they know it.

The goal of Computer Science is to build something that will last at least until we've finished building it.

Working...