Forgot your password?
typodupeerror
Social Networks

Bursting the Filter Bubble 136

Posted by Unknown Lamer
from the everyone-else-is-wrong dept.
Jah-Wren Ryel writes with news that a few CS folks are working on a way to present opposing viewpoints without angering the reader. From the article: "Computer scientists have discovered a way to number-crunch an individual's own preferences to recommend content from others with opposing views. The goal? To burst the 'filter bubble' that surrounds us with people we like and content that we agree with. A recent example of the filter bubble at work: Two people who googled the term 'BP.' One received links to investment news about BP while the other received links to the Deepwater Horizon oil spill, presumably as a result of some recommendation algorithm." From the paper's abstract: "We found that recommending topically relevant content from authors with opposite views in a baseline interface had a negative emotional effect. We saw that our organic visualization design reverts that effect. We also observed significant individual differences linked to evaluation of recommendations. Our results suggest that organic visualization may revert the negative effects of providing potentially sensitive content."
This discussion has been archived. No new comments can be posted.

Bursting the Filter Bubble

Comments Filter:
  • by Anonymous Coward on Monday December 02, 2013 @10:21PM (#45580831)

    i'm a generalist, i work in a few fields, including EE and CS - my colleague is pure CS

    we're trying to have a conversation about a topic (distributed clocks) and based on our histories
    we get entirely different search results, completely non-overlapping. his are general distributed
    systems results and mine are narrowly turned to sensor networks

    i had to ask him to make me a bibliography because I got sent into an entirely different
    alleyway of the literature

    thanks google

    • by sg_oneill (159032) on Monday December 02, 2013 @10:29PM (#45580859)

      Its even more problematic in areas like climate change where a large portion of the population appears unable to distinguish laymans commentary from actual research by climate scientists. If people spend a lot of time looking at conspiracy theory , creationist, or other similarly themed stuff on the net, google throws lots of denial sites at them, whereas people who have more analyical interests are more likely to get articles from science sites. The problem here is that folks with the conspiracy bent end up having no way to find information that might clear up their confusion if all they are getting is wattsup or alex jones or whatever. This just feeds the confirmation biases, and thats proving really harmful to science education right now.

      • by Burz (138833) on Monday December 02, 2013 @10:56PM (#45580929) Journal

        Most people don't buy into climate conspiracy theory. IMHO, this new method is more likely to be employed by paid Public Relations types to blunt pressure calling for social and ecological responsibility. If they can target unhinged conspiracies as "bubbles", they can preferentially target informed progressives (or any online community) to serve the interests of big business.

        I wouldn't trust the advertising business to be even-handed with acquired psychological tools.

      • by sribe (304414) on Monday December 02, 2013 @11:01PM (#45580951)

        The problem here is that folks with the conspiracy bent end up having no way to find information that might clear up their confusion if all they are getting is wattsup or alex jones or whatever.

        Your point is well-reasoned. But, unfortunately, I think you are starting from a false premise because you simply do not understand how delusions work.

      • Good point.
      • Erm, if you're only exposed to "concensus" views and around the edges you get sympathetic nutcases like Lewandowsky claiming anyone who disagrees is a fantasist with a mental problem (oh, the sweet irony), aren't you in exactly the position you are concerned about?

      • The problem here is that folks with the conspiracy bent end up having no way to find information that might clear up their confusion if all they are getting is wattsup or alex jones or whatever.

        This is the first time I've seen anyone try to claim that Anthonny Watt's blog was the stuff of conspiracy theory. It's a clever technique. If we can associate anyone that questions the IPCC with the likes of Alex Jones and David Icke, we can avoid having to address any questions they have at all.

    • by esldude (1157749)
      Yeah, realized this problem a few months back. I and a friend while on the phone typed absolutely identical search terms in. We got entirely different results. Often even highly different numbers of results. Like I returned 81,000 search results once and he got over a 400,000 for the same term. What we also found was if we kept typing in the same search terms somewhere around a dozen or more times we then started getting the same results. Sort of odd behavior. I thought. This using google of course.
    • by SnowZero (92219) on Tuesday December 03, 2013 @04:41AM (#45581999)

      It is exceedingly unlikely that the results don't overlap after the first few, but if you can produce a copy of the two sets of results, I will forward them to someone on the Google Search team for debugging.

      People hugely overestimate the effect of personalization -- it is a ranking tweak not a complete change to the search engine. It does not make economic sense to have personalized whole-web indexes.

      Btw, if you don't like personalization ever, it is pretty easy to turn off:
          https://support.google.com/accounts/answer/54048?hl=en [google.com]
      Just remove web history and uncheck private results.

      • by Gr8Apes (679165)

        Based on what's being responded to this post, it doesn't seem an isolated case. Granted, there's no proof, but given what we've all seen, it appears to be at least somewhat true.

        The simple solution is simply using a "trash" browser instance that you can completely clear all cookies and local data from, you can avoid the personalization almost entirely, at least for a short while.

    • Does Google adapt based on IP even if you never log into their "services"? I never log in and don't see any results tailored to me, but now I'm wondering whether that's actually true.

      Regarding the paper: Don't bother reading it, it's superficial CS stuff with no proper background theory, they don't even model preferences right. :-(

      • I never log in and don't think I see any results tailored to me, but now I'm wondering whether that's actually true.

        You'll never really know will you? But I wonder how they handle disparate groups sharing a single IP in that case.

        • I never log in and don't think I see any results tailored to me, but now I'm wondering whether that's actually true.

          You'll never really know will you?

          Isn't that testable? Set up two identical browsers on two geographically close machines, make them perform a series of differing and clearly biased searches (e.g. about a political topic or shopping preference) and compare the results of one and the same query from both machines after a few days or weeks.

          I was hoping somebody did that already...

    • One should be very concerned when person A gets one set of results, and person B gets another set of results; both using the same search criteria.

      I never thought that Myopia was communicable.
    • by Burz (138833)

      You could open an "incognito" or Private browser window for the search, then your cookies would not be visible and your history would not be applied.

  • Critical thinking (Score:5, Insightful)

    by nickmalthus (972450) on Monday December 02, 2013 @10:22PM (#45580835)
    "I know you won't believe me, but the highest form of Human Excellence is to question oneself and others." - Socrates

    It is good to see someone researching ways to combat group think with technology.
    • horsedrinkwater (Score:5, Insightful)

      by globaljustin (574257) <justinglobal AT gmail DOT com> on Monday December 02, 2013 @11:06PM (#45580971) Homepage Journal

      I'm with you in wanting to combat "groupthink" but I don't think the technology in TFA will do it

      First, who 'scores' the viewpoints? how is one organization weighted against another? ex: Fox News should be in the tabloid/nonsense news category but because Fox is kind of 'grandfathered' in as the 4th national network they are considered 'mainstream'

      does this mean a person who goes to motherjones.com alot would get Fox News in this system? who determines that?

      i would consider Fox News a 'lower' viewpoint...different sure, but not in any value added way....ignorance isn't an "opposing viewpoint"

      2nd, is this going to be an "add-on"? Is the goal to get Google, etc to use it by default?

      because people would ignore this tech for the same reason they don't bother seeking out differing viewpoints!!!

      unless you force it on them the people who need it won't do it!!!

      3rd, if forced upon them, people will inevitably train themselves to ignore the 'suggested alternate viewpoint' box just like they train themselves to ignore Google.com's "sponsored results" or tune out a commercial

      To me, this is an example of why academics fail in public policy. They look at a problem and see human opposition as something uncategorizable so instead of understanding that **the problem isn't that people don't get opposing viewpoints...the problem is they willfully choose not to listen**

      as they say, "You can lead a horse to water but you can't force them to drink"

      this is like holding the horse's face in a water fountain

      • I'm sure that sometimes people choose not to read opposing viewpoints. On the other hand, I very much enjoy the opposing viewpoint when it's presented in the style I prefer - with logic, fact based, and backed up with details like relevant numbers. Very often, it's not so much the opposing conclusion that turns me off, but the illogical, purely emotional and often sarcastic presentation.

        In the post I'm replying to, for example, I enjoyed the second part, discussing possible reasons, but wouldn't have clic

      • Fox News should be in the tabloid/nonsense news category but because Fox is kind of 'grandfathered' in as the 4th national network they are considered 'mainstream'

        No, it's mainstream because more people watch it for news than anything.

        You may disagree with the ideological bent but you seem to be confusing heavily partisan shows that are on Fox News channel, with the actual news coverage - of which there is a lot.

        Fox news covered all of the major news stories as well as any other channel - like trains going

          • How is mainstream, by definition, not the most popular and heavily watched news channels? That is what the "main" group of news watchers are watching - so how can it not be Mainstream?

            What is nonsense is ignoring the definition of a word because it results in something you disagree with.

    • by TubeSteak (669689)

      Don't confuse group think and tribalism.
      The main difference is that group think is a problem for leaders, while tribalism is what motivates all the followers.

      You can't combat tribalism nearly so easily as group think,
      since the group identification is more than just a matter of facts.
      Otherwise, the Red Sox wouldn't have any fans.

    • by smpoole7 (1467717) on Tuesday December 03, 2013 @12:20AM (#45581275) Homepage

      > "I know you won't believe me, but the highest form of Human Excellence is to question oneself and others." - Socrates

      > It is good to see someone researching ways to combat group think with technology.

      But that always, always starts with the guy in the mirror. First, get your own mind right (as Socrates says).

      Next, most people listen to those friends whom they respect. You can challenge them to examine alternative points of view. The only thing I would ask (of everyone) is that you respect people who look at things as honestly as they know how, but reach a different conclusion from you. That's part of the human condition. The name-calling and "group think," as you call it, stops when we decide that it will stop.

      I lean conservative/libertarian in philosophy, but I avoid polemics from all sides. My morning ritual nowadays consists of first checking the weather (because of my job), then heading to Real Clear Politics (www.realclearpolitics.com) to get a diversity of opinion, from Ezra Klein and Robert Kuttner to George Will and Mark Steyn. I also love a good (read: FRIENDLY) debate. If I see name-calling on either side, I lose interest in a hurry.

      But have a friendly discussion with your friends. The old saying goes, "don't discuss politics or religion," but I say the opposite. If you show them respect, they'll learn to respect you, and in turn, they'll learn to respect opposing points of view. You might even learn a few things.

      I certainly have. :)

  • We found that recommending topically relevant content from authors with opposite views in a baseline interface had a negative emotional effect. We saw that our organic visualization design reverts that effect. We also observed significant individual differences linked to evaluation of recommendations. Our results suggest that organic visualization may revert the negative effects of providing potentially sensitive content.

    Reads like somebody trying to write in English, and utterly failing.

    • by ackthpt (218170)

      We found that recommending topically relevant content from authors with opposite views in a baseline interface had a negative emotional effect. We saw that our organic visualization design reverts that effect. We also observed significant individual differences linked to evaluation of recommendations. Our results suggest that organic visualization may revert the negative effects of providing potentially sensitive content.

      Reads like somebody trying to write in English, and utterly failing.

      Sounds like the sort who want to find only what they've already found.

    • by Arker (91948)

      I couldnt say it fails to be English. However after perusing every link I am left without any information about what this 'organic visualization' thing actually is, and how it's supposed to work.

      Leaving me only to comment more generally that it is not the search engines job to read the users mind, and shouldnt be, if only because it is impossible to do that with any amount of computing power. It's one of those disastrously attractive ideas that may take us another decade to finally start getting over.

    • by Animats (122034) on Tuesday December 03, 2013 @12:07AM (#45581209) Homepage

      The "organic visualization" thing and its jargon are described in this thesis [benfry.com] done at the MIT Media Lab. This is what happens when postmodernists try to improve on Tufte. [edwardtufte.com] Some of it is pretentious bullshit. But there may be the genesis of some new phone apps in there.

      Here's a good, but unrelated, example of "organic visualization": BitListen [bitlisten.com] This is a little HTML5/JavaScript page which depicts transactions on the Bitcoin block chain. An older example is Muckety [muckety.com]. This can be done well, but most attempts in this direction are duds.

  • n/t
  • by mattr (78516) <mattr&telebody,com> on Monday December 02, 2013 @10:52PM (#45580923) Homepage Journal

    I only skimmed the paper briefly but it is interesting in that:
    - User clicks a wordcloud keyword/hashtag that draws lines from it to multiple florets (individual nacelle-like microflowers in a sunflower head), each of which represents a tweet in recent portion of a feed.
    - Repudiates the idea of filtering to meet viewer expectations so everyone can see the same content.
    - A cuteness factor (or what they say is "organic" being like a flower) apparently reduces gut reaction to tweets you do not agree with
    - Viewer is able to actively pick tweets to read. Presumably as the sunflower head image is mathematically generated and each floret's color could be tweaked to match a positive/negative sentiment score, allowing the user to pick only items that agree/disagree with them but to do so consciously.

    This last point would seem to be ideal and I'd like to see slashdot include something more than the slider ("read only above this score"), particularly for a topic that has over say 500 or 800 replies. How about a data visualization that shows all the posts/threads for an article and lets the user select based on where in this chart a post is? At the very list, something 2-dimensional not 1-dimensional.

  • How about... (Score:4, Insightful)

    by ackthpt (218170) on Monday December 02, 2013 @10:59PM (#45580945) Homepage Journal

    We quit this crap of trying to target things to audiences and get back to the good old days of yore when we went out and found things to fascinate, inform and enrich ourselves rather than suffering pigeon-holing. Honestly, I think farcebook, amazon and others have it completely wrong. I'm bored by the same ol - same ol. I'm an explorer and love to wander and see new things. Keep showing me what i've already seen or already bought and I'm losing attention.

    • Re:How about... (Score:4, Insightful)

      by JanneM (7445) on Monday December 02, 2013 @11:19PM (#45581037) Homepage

      What "good old days" where those? When you read the newspaper that conformed to your political viewpoint; the weekly magazine that covered any world events only as far as it affected you and others like you; watched only the TV shows that reinforced what you thought you already knew and believed?

      • Re:How about... (Score:5, Insightful)

        by ackthpt (218170) on Monday December 02, 2013 @11:31PM (#45581101) Homepage Journal

        What "good old days" where those? When you read the newspaper that conformed to your political viewpoint; the weekly magazine that covered any world events only as far as it affected you and others like you; watched only the TV shows that reinforced what you thought you already knew and believed?

        No.

        The internet before all this tracking of metrics and trying to anticipate what I'd like to see more of. I don't know what I want to see next, but I generally don't revisit the same old thing. After I bought a new camera is not the time to keep showing me camera stuff. When I looked up something on ebay to see what I might get for it, they keep trying to interest me in it over a year later - I don't buy everything I look at and there's no "I'm just trying to get an estimate of what I might get from a suck^H^H^H^Hbuyer so piss off and don't try to waggle it under my eyes for the next twelve bloody months" tick-box.

        Just anecdotal, but the things facebook seems to track and then keep showing me have about 95% odds of not being of interest at all, the remaining 5% I wouldn't click on a link on there anyway or it's only tangentially relevant to something I was posting about.

        After a while I just tune stuff out.

        • by AHuxley (892839)
          At some point price customisation or "dynamic" pricing must have become very useful and widespread.
          http://edition.cnn.com/2005/LAW/06/24/ramasastry.website.prices/ [cnn.com]
          Your search habits fed back into a price just for 'you' at that moment on a sales site.
          • by ackthpt (218170)

            At some point price customisation or "dynamic" pricing must have become very useful and widespread.

            http://edition.cnn.com/2005/LAW/06/24/ramasastry.website.prices/ [cnn.com]

            Your search habits fed back into a price just for 'you' at that moment on a sales site.

            Alas, that was what I was interested in yesterday or at some time in the past. It's like being in a discussion with someone who keeps bringing up a point that you thought was resolved a while ago, but they think needs more exposure. Imagine how you'd enjoy such a discussion.

          • Are we to a point where instead of "ad words" Google has come to the realization that a company can PAY not to have bad gossip said about them?

            This is so much like TV News and commercial advertising. Great. Now the internet is as broken as the TV.

            [insert frowny face here]

        • Just because I might LIKE to confirm my world view -- is it in societies best interest to automate an "eco chamber" on the internet?

          The rights of privacy only extend to our Government -- there's still this gray area with services and corporations. I don't think there is any clearer scenario to say; "we've gone to far." It's one thing to optimize a search -- it's another to play a digital psychologist and give different results to different people based on their profile. Down that path leads people who don't

  • Push vs Pull (Score:5, Insightful)

    by quantaman (517394) on Monday December 02, 2013 @11:24PM (#45581065)

    If you search for X, and get confronted with an adversarial opinion, the contrary information is being pushed at you which is threatening and probably responsible for the negative emotional reaction.

    If you search for X, see where the adversarial opinions are, but don't actually have to see them when you want to, that's more a pull mechanism and you feel much less threatened as a result.

    From what I can tell glancing at the paper their system is very much a pull mechanism which probably lowers the negative response.

  • Wait what? (Score:4, Funny)

    by Charliemopps (1157495) on Monday December 02, 2013 @11:55PM (#45581175)

    Two people who googled the term 'BP.' One received links to investment news about BP while the other received links to the Deepwater Horizon oil spill, presumably as a result of some recommendation algorithm.

    So I tried this, I wanted to see if the Internet thought I was a democrat or a republican but it just came up with a bunch of links to "Big Penises" I was outraged, and after 15 to 30 min or so of confirming the content I switched search engines. The algorithms are clearly out of control!

    • Two people who googled the term 'BP.' One received links to investment news about BP while the other received links to the Deepwater Horizon oil spill, presumably as a result of some recommendation algorithm.

      So I tried this, I wanted to see if the Internet thought I was a democrat or a republican but it just came up with a bunch of links to "Big Penises" I was outraged, and after 15 to 30 min or so of confirming the content I switched search engines. The algorithms are clearly out of control!

      I just tried this and got... financial news, info about the oil spill on both sides of the coin (i.e. How BP got screwed, What BP doesn't want you to know), and local gas stations. I guess this means that I'm interested in the stock market (which I am, though I don't own BP stock), I'm open to understanding the facts on both sides (i.e. a balanced viewpoint), and I recently searched for gas stations (which I did, but for the local Shell stations not BP).

  • Wiping all your cookies and history in your browser might help a bit, but probably not if you're using Chrome or logged into Google.

    Also, the Startpage search engine claims to use Google, but anonymously.

  • Using a tool means that you are already aware of the problem. Lets suffice to say: most people are not, and also may even feel content in their bubble. And that's the real problem.

  • So this algorithm can figure out what your political viewpoint is, as well as the viewpoint of a news article.

    This could be used to show you articles which contradict your viewpoint - or articles which reinforce your viewpoint.

    I think the latter is more likely in practice. It's more profitable.

  • I tried various searches in a normal browser window and an incognito window. There was really very little difference. Searching for "Laptop" swapped the order of amazon and curries. Searching for Islam had three appeasement sites at the top in my search and five in the incognito search. Searching for "BP" was exactly the same.
  • by supercrisp (936036) on Tuesday December 03, 2013 @07:29AM (#45582419)
    I keep reading about this bubble, but I don't experience in my daily life. I am by political inclination pretty far to the left, but I run into plenty of right-wing opinions, from the libertarians on Slashdot to the Tea Party people on Facebook. I interact with moderate Republicans at work and extreme (God needs to cleanse this nation! Gold Standard!) Republicans in my neighborhood. I have no sense that there's a bubble. I sometimes wish there was a bubble that could filter out all the idiots. Some of the best days of my life were spent hanging out with people of varied and conflicting views who were all intelligent and capable of mutual respect and civility. I'd love a bubble like that. But, again, I don't see any damn bubble in my daily life. Why's it getting broadcast so much? Cui bono?
    • I think you misunderstand the bubble. Think advertisment, news article filtering, search results... You are describing self-subscription to things outside your viewpoints. The bubble doesn't keep you from hearing those viewpoints when you choose to go get them; it just subtly confirms its projection of your own viewpoints when you aren't specifically looking for something conservative. For instance, if I'm logged in to Yahoo, it incorrectly labels me Republican and will give me articles slanted to conse
    • When you interact with these people do you ever notice that they've got certain "facts" that you do not?

      The "bubble" is of information that is unique to the perspective -- or reinforcing that perspective. It can also be called an echo chamber where you hear back what you preach.

      The fact that you don't "see" a bubble is due to you not being cocooned in your own rhetoric. Meeting people who disagree and not getting horribly bothered is a sign of good mental health.

      / begin jokes here ////
      Personally, I look for

  • You could learn a bit of tolerance and use a search engine that doesn't try to profile you like https://ixquick.com./ [ixquick.com.]

    Rosie

  • Not with the general population, the prejudice is inside them, and even if you slip in an opposing view once they notice that's what's been done you've lost them again.
    I would say it's almost impossible to change average adult minds and opinions on things they consider important by having them read articles with opposing views.

  • Seriously, what the hell?

    I understand how people can live in an echo chamber and suffer from monoculture. And targeted ads and recommended searches and all that jazz can contribute.

    But, seriously? Just log out of google. Run your search. If they don't know who you are, you get an unfiltered search.

    Is that so hard?

  • I didn't even know that google is doing this. One more reason to kill their cookies often (or not accept them in the first place...)
  • I mean, I occasionally read stuff from the other side,,, but a fair bit of it ranges from Michelle Bachman (R-wacko)'s la-la land pronouncements, to a guy on a mailing list I'm on, who's sure that with the Fed printing money, it's devaluating as we type, and oh, yes, we live in a soclaist America....

    How does this "bubble-bursting filter" screen out the total crazies, to get *reasonable* contrasting information?

                      mark

  • They're only recommending things be filtered differently, not that they should not be filtered at all. Useless. I prefer to use an anonymizing search engine, and browsers that do a better job of it would also be a help. The problem is to remove any ability for a site to know anything about the user at all, not to just be "nicer" about how they filter based on what they know.
  • The "Filter Bubble" is in part the product of too much information and a very poor design for public discussion, the blog, which is intentional. Even Slash Dot has the core flaw even though it is a forum, technically. If there is a filter bubble at all, it is in part created by the need for moderation of blogs and the lack of a good mechanism to handle topic drift, sub-threads, and the way to create an interest, even compelling one, for opposing or different views.

    I draw on my recall of how on the USENET

Ever notice that even the busiest people are never too busy to tell you just how busy they are?

Working...