Slashdot Mirror
Facebook Patents Inferring Income of Users
theodp writes "Among the patents granted to Facebook this week by the USPTO is one for Inferring Household Income for Users of a Social Networking System. 'For example,' Facebook explains, 'an assumption might be made about a user that reads CNN.com and nytimes.com every day that the user is in a higher income bracket than another user that only reads TMZ.com and PerezHilton.com on the theory that a user who reads newspapers might be assumed to make more money than a user who only reads celebrity gossip blogs.' Advertisements such as those for travel packages, cars, and home mortgages, Facebook adds, 'are targeted to users based on income bracket,' which might also be inferred by 'gathering and analyzing different types of information about a user's geographic location.' Hey, what could go wrong?"
Browse anonymously
This is why I use EasyPrivacy list in adblock plus to keep Facebook from getting that info. They know you read a page if it has a "Like" button on it.
FB, you shouldn't stop there! just patent all statistical research!
It doesn't really matter if the algorithm is wrong for an individual, as long as it it generally correct for the population.
The reason Facebook has any advertising income, and therefore value as a company, is that it has the ability to provide very directed advertising.
If you want to target people who read cnn.com and nytimes.com, why not just advertise there like you always could.
Another drop in the ocean of stupidity. How long until the politicians realise this slows the whole economy down? I'm not holding my breath....;-(
They do know everyone selling data + advertising already does this, right? This is a VERY obvious use of aggregated data.
I declare Shenanigans!
Shenanigans on Facebook!
Shenanigans on the USPTO!
They are trying to scam us now and it needs to be stopped Officer Barbrady!
Jokes on you FB... I read all of those websites every day.
Whenever someone uses a correlation statistic? What about when someone uses a set of data to infer something about other populations?
It may sound an awful lot like they patented statistics, correlations, and sampling, but it's different 'cause it's on a social network. Totally different.
I live in constant fear of the Coming of the Red Spiders.
I post 5 times a year and only sign on via VPNs, never my real IP..
My low, low, government salary.
Can I file a claim that (oh sorry, "what is claimed is that...") the lawyers filing patents for Facebook make even more money on average than readers of CNN.com of nytimes.com?
Coming soon to slashdot, targeted Linux distribution adverts. Can slashdot infer what distro you are into based on your browsing history? this AC thinks yes.
Not sure how redlining applies here, seems like much of a stretch from the 1934 National Housing Act* proscribing to the FHA to create redlining three decades before it got that name, to displaying advertisements based on viewing habits.
*See first paragraph in the History section of submitters link on redlining.
Time Bomber the Book coming soon.
what's next? inferring men genitalia size? i can't believe the general public just doesn't care about their privacy being so violated in exchange for a beef stew. But then i heard that having no facebook profile is an indicator for being a psychopath, so who knows anymore.
A system for applying common stereotypes to circumvent privacy restrictions that is not covered by the DMCA...
Add this to an ever growing list of reasons to NEVER use Facebook.
In a post-capitalist economy, this is the 1% new power version of racial profiling. Think of it as a form of denial of services. You'll never see what the 1% does much less enjoy. Knowledge is power and denying information denies access to all but the 1% who matter.
http://www.tivocommunity.com/tivo-vb/archive/index.php/t-287253.html
If enough users launch it, it will completely mess up all these statistical correlations and eventually provide anonymity by increasing the noise.
sed -e 's/Chuck Norris/Rajnikant/g' joke > fact
If algorithms can be patented, then sure. If FB is using a unique algorithm to infer income, it might be granted (that I think patenting mathematics is absurd is irrelevant - if you believe your algorithm is so great, keep it a secret. Application of mathematics to one area shouldn't be patentable). I'd be surprised if Amazon doesn't look at your shopping history and suggest products in your price range. If I never bought anything over $25, why should they show me a product costing over $10,000?
On the other hand, what does this have to do with redlining? My outrage that statistics is being patented has nothing to do with the fact that FB should be allowed to show whatever ads to whomever they please. They are not a government organization (and haven't taken taxpayer money) that shouldn't be allowed to discriminate between consumers.
corn is a powerful consumer classification that segments the UK population. By analysing demographic data, social factors, population and consumer behaviour, it provides precise information and an understanding of different types of people. Acorn provides valuable consumer insight helping you target, acquire and develop profitable customer relationships and improve service delivery.
This is just another case of adding "... on a computer" or "... over wifi" to something that's already an established practice to gain a patent.
Doesn't bother me. I said Fuck FaceBook years ago.
There is an browser addon called Disconnect that blocks your browser from loading most if not all tracking resources, this includes the social media buttons used by Facebook to track your browsing. I put it onto every browser I come into contact with that supports it. https://disconnect.me/
Parent jokes, but it's really eerie the way the marketing statisticians come up with this stuff.
It brings up a whole philosophical question: if we can be pigeon holed so easily, are we really as free thinking as we think we are?
I think it's more likely that they can infer what you make BY WHAT CURRENT JOB YOU LIST. Someone listed as working as a lab tech somewhere is obviously going to make more than someone listed as working at McDonald's.
Pretty, pretty low.
Here's an idea: "Inferring Employment Status From Branded Apparel". If we see someone wearing a Facebook T-Shirt, we can infer that they a) work for Facebook, b) know someone who works for Facebook, c) attended a Facebook event, or d) knows someone who falls into one of the previous categories. What a mind-blowing invention! What a truly original and wholly non-obvious contribution to society!
It's not inferring, it's *implying* .. In the case of facebook, the more they use, the more they lose.
There really is a sense of classism up there isn't it? I read the sites mentioned daily, and I make 50k a year. This is like assuming someone's black due to the number of fried chicken ads they clicked.
Reflecting back to the "SCOTUS reviewing software patents" article on slashdot yesterday, it sort of makes me think how this "invention" would go down in real life, without a computer or the internet.....perhaps something like this:
*Joe Shopper enters Costco*
Costco Greeter: Hello! Would you mind telling me where you came from before entering our store and what kind of car you drive?
Joe Shopper: Why?
Costco Greeter: We would like to use that information to take a guess on how much money you make?
Joe Shopper: Why don't you just ask me?
Costco Greeter: Oh... hmm.. ok... how much money do you make?
Joe Shopper: GFY.
*Jack Shopper enters Walmart*
Walmart Greeter: Hello! Would you mind telling me where you came from before entering our store and what kind of car you drive?
Jack Shopper: *flees*
Walmart Greeter: *crawls into fetal position and cries out the remaining 11 hours*
So like... in real life, without computers and the internet, the task would be hard to complete, yeah? Not because of the technical details of the task [you ask, they tell], but more due to the individuals privacy. The first clue to facebook should be that they are having to use a surreptitious method to make an assumption rather than directly asking you what your household income is.
That's how they got Al Capone...
Didn't know you could patent that.
Facebook's advertising is all about micro targeting based on a user's behavior, open to all with the smarts to read it via the Open Graph API.
Advertising (TV, Radio, Banners, Internet) is sold on an open market bidding system. You bid for impressions (CPI) and clicks (CPC). Coveted demographics - such as 16-20 year old females, or wealthy folks have very high bid rates. So being able to infer people's income makes good business sense.
Facebook has a good model, as you bid for placement based on age, location, brands they fan, web pages they like, number of friends, etc. as opposed to blindly putting advertising out there and hoping the right folks read it. Many local small businesses would be out of business if it wasn't for this. Those are the folks who are creating most of the new jobs. Google, on the other hand, uses your search behavior, and tracking cookies left by banner ads + stuff you look at. If you use free or paid gmail the emails you send and receive come into play. That's why it's free, of course. The other "free" email services do the same thing.
There's a database called PRIZM that's in widespread use that determines relative income based on zip code - PRIZM has been around since the direct mail days. My guess is that Facebook either does not want to pay for this, or because it doesn't have user's zip codes, only cities, it thinks it can do better based on data it has in the social stream.
Being able to predict the precise moment when purchase intent is realized... is the holy grail of digital marketing right now, with billions of dollars being spent on the problem. Every piece of data about your online behavior is being analyzed against what you actually bought (and if you're on a mobile device, where you were at the time). It is only a matter of time before this is perfected - e.g. the moment you realize you are hungry an ad for nearest food joint of your preference will appear in front of you... This trend is as unstoppable as dawn, and if you're smart, you're trying to patent every single idea that might remotely be involved in this.
If you really think you can protect yourself from companies analyzing your behavior, you're dreaming. Being able to deliver the ad to you at the precise moment of purchase intent will save businesses large and small billions and billions of dollars... And the first company to make that work, will be profitable beyond belief.
Murphy was an optimist
Another confirmation that our idea of the internet has devolved in the hands of entrepeneurs.
Agreed. There is probably plenty of prior art, but one would be crazy to challenge the patent because:
Congratulations, Facebook, you are a patent troll
http://yetanotherpoliticalrant.blogspot.com
Isn't this exactly what marketing research companies have done before. A quick web search says :
corn is a powerful consumer classification that segments the UK population. By analysing demographic data, social factors, population and consumer behaviour, it provides precise information and an understanding of different types of people. Acorn provides valuable consumer insight helping you target, acquire and develop profitable customer relationships and improve service delivery.
This is just another case of adding "... on a computer" or "... over wifi" to something that's already an established practice to gain a patent.
No, this is another case of not reading the claims. The patent claims go into confidence metrics and applying advertising criteria based on those metrics. Now, maybe there's other prior art out there that teaches that element, but your link is the equivalent of saying "Tesla got a patent on the power train in the Model S? But isn't that really just a Ford Model T adding '... with a battery'?"
A very informative site you update here I really found a lot of information about many objects. I am trying to update my blog http://www.fashionvela.com/ which is regarding fashion and trying to find new things through browsing many sites in a day
How about them Knicks?
This is a ridiculous patent and should be invalidated. As others have said, this is correlation. Nothing patentable here at all.
I taught data mining in college. This is a standard example of relating attributes to income. It is not novel.
This is one of two things: A multiple regression (or any general linear model for predicting the dollar amount) or a multinomial logistic regression (any generalized linear model for predicting income category). Both of these are just maths (yes, we could do something more complex, with a genereal additive models, or MARS, et cet), but all they did is provide large amounts of training data. Given the same data I or any statistician do the exact same thing.
WHAT THE FUCK. I don't normally give useless responses.. but WHAT THE FUCK. PATENT REFORM NOW.
My Town is unusual in that we have an equal spread of folks in each quintile. That is, we have a few legit plutocrats and a few folks "on scholarship" for school lunch, and everyone in between. Unless you are the NSA or IRS (or both) income is a tricky thing to guess.
You don't have to unblock Facebook to use most comment sections. More of the major new sites are using either Disqus or a site-specific instance of LiveFyre than are using Facebook Comments as their enhanced commenting platform. USA Today is probably the biggest site using Facebook Comments. A lot of local news stations and small-town papers have moved to Facebook Comments. Lots of blogs and special interest websites now use Disqus to get into that cross-web "discoverability" of their sites by being on the same comment platform as CNN, The Atlantic, etc. Some sites still use Intense Debate, though it's dropped off bigtime. Wonkette probably the biggest political commentary site still using it, some blogs, some small news sites. (Intense Debate had the "early mover disadvantage" - LiveFyre and Disqus are just much better.)
Even for the Facebook Comments-powered sites, you don't have to unblock Facebook globally, if you use the right tool.
Problem: You don't want to be tracked by Facebook all over creation, but you do want to be able to comment on the majority of sites. Including, if they use Facebook comments, those sites.
Solution: Use Ghostery (and I'm specifically recommending Ghostery, not alternatives like Disconnect; I explain why further in) with its fine granularity of global and site-specific blocking.
1. Turn off GhostRank, so you're not telling Evidon (Ghostery) who you're going to. It's off by default so they're being good guys.
2. Turn on auto-update and auto-block new elements.
3. Block everything. (It's just easier to start from blocking everything. 3 after 2 because sometimes first-use leaves stuff unblocked)
4. If you're a regular commenter and comment reader at major sites, unblock the "3pes" (Third Party Tracking Elements) for:
Disqus
LiveFyre
Intense Debate
If using the Firefox version of Ghostery, there's a Cookie tab. Repeat steps 3 and 4 on the Cookie tab.
Disqus and Intense Debate have cookies on their list, too, LiveFyre currently does not.)
5. Save (one save covers all the tab settings you've jumped between.)
Do not unblock Facebook or anything with Facebook in it here at the global level. You don't want Facebook knowing every site you've been at that has a Like or Follow button or a Facebook Social Reader app, just the ones you intend to actually read Facebook-powered comments at.
The last several versions of Ghostery for Firefox, and the most recent version for Chrome finally, have per-site per-tracker disabling. So go to the site where you can't see the comments. Click the Ghostery toolbar icon to see the list of trackers blocked. Don't whitelist the whole site. Next to each active tracker, Ghostery has a slide switch. You can unblock Facebook Connect or Facebook Social Graph or whatever you need, just for that site, then reload.
It may well turn iterative. For Facebook comments it certainly will. On USA Today, for example, if you click the little dialog bubble icon on the left panel from the story (which is their comment icon), Ghostery will increment by at least one more tracker, USAtoday didn't load the FB stuff till then. Unblock that and reload, you still won't get the comments. By unblocking Facebook Connect, now it could load Facebook Social Plugins. Now unblock that. Rinse and repeat.
I'm a pretty avid Disqus commenter and have it on all my and my clients' sites, so I leave it unblocked globally. But you could do the same with that, if you only want it to work at certain sites and don't want it knowing you're there at other Disqus-powered sites.
One thing I've found on a lot of sites - even with Disqus (or LiveFyre) unblocked, the site's JavaScript that in turn triggers the Disqus or LiveFyre plugin, won't fire unless you unblock something else. And sometimes that "something else" isn't particularly "safe" for folks who don't want any adverts or cross-web trackers. Omniture from Adobe's advertisin
... has been inferring a potential customer/candidate's value. Forever. How can one particular "inventor" possibly claim that it's novel and unobvious?
I just have to wonder what anyone who watches my shopping hairs would think.
In fact I buya lot of stuff for unusual reasons, or because my friends ask me to buy it for them for various reasons.
I've bought medicine for genital crabs/lice, even though I've never had them (they were all out of the feline kind and the pharmacist told me it would work - he was right! But it did cost more.)
I've bought pregnancy tests, tampons, and yeast infection medicine for my female friend who were too shy to buy it in person.
I've bought hair loss treatments, even though I'm not at that point yet. I've bought baby clothes and other things not needed in my household as gifts, etc. Etc. Etc.
Who knows what someone would think if they judged me by my shopping habits!
Is there anything I can do to convince Facebook and other social media sites that I will never ever respond to advertising, and therefore they stop filling my feed with crap?
Thought not.