Slashdot Mirror
Real-World Outcomes Predicted Using Social Media
Hugh Pickens writes "Kevin Kelly writes that researchers at the Social Computing Lab at HP Labs in Palo Alto have found that social media content can predict real world outcomes. In their study, the researchers built a model that used chatter from Twitter to predict accurately the box-office revenues of upcoming movies weeks before the movies were released. When the sentiment of the tweet was factored in (how favorable it was toward the new movie), the prediction was even more exact. To quantify the sentiments in 3 million tweets, the team used anonymous workers from Amazon's Mechanical Turk to rate a sample of tweets, and then trained an algorithmic classifier to derive a rating for the rest. But predicting box office receipts may be only the beginning. 'This method can be extended to a large panoply of topics [PDF], ranging from the future rating of products to agenda setting and election outcomes,' the researchers write. 'At a deeper level, this work shows how social media expresses a collective wisdom which, when properly tapped, can yield an extremely powerful and accurate indicator of future outcomes.'"
Predicting what a bunch of assholes who use twitter will do.
These social network predictions were already predicted by the late Mr. Asimov. ;)
The larger your sample size the more accurate your results tends to be. Fascinating.
Tweet incorrectly about the next election to screw with the people tracking this.
At a deeper level, this work shows how social media expresses a collective wisdom which, when properly tapped, can yield an extremely powerful and accurate indicator of future outcomes.
In other words - Never underestimate the power of stupid people in large groups.
A candidate makes a speech, and then the workers scrape SN data to see the response, so the candidate can further tailor the message for the next speech. However, do we really want to always be driven by public opinion?
The fact that the tweets predict the sales in advance of the movies release or people actually seeing the movie raises an interesting question. Is it the content of the movie or the "buzz" that really matters?
Statesman
Is this an early experiment in the development of psychohistory?
Hari Seldon would be proud.
It works until marketing departments at big companies start gaming it in a big way.
The truth is that all men having power ought to be mistrusted. James Madison
Wikipedia says:
The Delphi method is a systematic, interactive forecasting method which relies on a panel of experts.
Of course, in this case the "experts" are the movie-going public, who know more about their tastes in movies that anyone is Hollywood. The Delphi method depends on large panels, and n this case th researchers are using large panels indeed. Finally, the iteration is provided by the later tweeters reading earlier tweets before they post.
Nothing for 6-digit uids?
Putting the fact that increasing sample size does not necessarily increase the power of a predictor, you apparently didn't get the point of their method.
So method A was to simply "grep RamboIX" in these 3 million tweets. That alone already correlated to the box office outcome. However, that also catches messages like "RamboIX suxx, no way I'm going to see or even download this".
So method B was to use machine learning algorithms, combined with some initial work by human drones, to assign a degree of "positiveness" to each message about RamboIX.
While this has nothing to do with increasing sample size, it took the accuracy of the prediction to a whole new level.
I for one think this is a pretty great idea.
Collective wisdom? In today's popular culture? Surely there must be a better term than that.
Yes, 'the crowd' generally has a pretty good idea of what 'the crowd' is going to do.
If you ask me if I'm going to go see a movie or not than my answer is probably going to pretty accurately reflect what I'm actually going to do.
I'm not exactly sure why this is surprising? Marketers have been doing this for years. They announce products that haven't even hit the drawing board yet. If they get a good response in the form of inquires and other types of interest, they build the device. If they don't then it will just go away.
Persistent Volume manager for Kubernetes - https://github.com/dwimsey/openshift-pvmanager
Forget the mechanical turk. Just tell me the marketing budget of a movie and how many screens it's going to open on and I'll give you an estimate of the take. It's a pretty strong correlation on large-release movies.
Hollywood has worked so hard to remove the actual quality of the movie from the equation. Get the movie onto a lot of screens early and spend a lot on advertising. Get people in to see it on the first 3 days (Fri, Sat Sun, or Wed-Sun in some cases) before any info about the movie that you don't control (i.e. other than promos) gets out.
Many movies make around half their total theatrical take in the first weekend of release.
http://lkml.org/lkml/2005/8/20/95
When it happens??? It's been happening from day 1. They're just so good at it that you haven't noticed.
I'm out of my mind right now, but feel free to leave a message.....
It works until marketing departments at big companies start gaming it in a big way.
No need to game the system when you've already gamed the users! How do you think all these twitterers know to talk about movies before they're released?
The quote in the article is such crap:
At a deeper level, this work shows how social media expresses a collective wisdom which, when properly tapped, can yield an extremely powerful and accurate indicator of future outcomes.
No it doesn't show that at all. It shows that what is popular in twitter is popular in the real world. In other words, it shows that twitter is close enough to a representative sample of the general population for many practical purposes. That is all. It doesn't have anything to do with collective wisdom, nor does it help you predict any outcome unless it is primarily dependent on popular opinion.
Wow. It's like psychohistory - uh, wait - isn't this really just doing a poll? Hasn't this already been done for decades? Just poll a small number of people in the general population, and you can "predict" all kinds of things - like elections before they happen? Sigh. The only difference between this and regular polls is that this is less scientific (since they make no effort to find a random selection of people from the population). It's probably a little better than online polls (probably less manipulation) and a little worse than scientifically-designed polls.