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?"
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.
The patent specifically suggests using inferred income for targeting mortgage offers, which the Wikipedia article notes has been a ripe area for abuse: "Reverse redlining occurs when a lender or insurer targets minority consumers, not to deny them loans or insurance, but rather to charge them more than could be charged to a comparable majority consumer whose business is more sought after"
Browse anonymously
Excellent; that puts you into a group of 0.001%. As a highly technically literate user you will have monthly household income of between 10,000 and 11,500 (95% chance) and are less likely to want pop star accessories. However, there is a 30% increase in the chance of you purchasing electronic gadgets. The correlation of your IP address with a slashdot reader decreases your chance of wanting to buy wedding accessories by over 99.72%.
Stay anon; please.
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.
"Post-capitalist"? We're reaching the zenith of everything Capitalism has sought to achieve. Record income disparities with unparalleled wealth for the super-wealthy, concentrating control over every aspect of society in the hands of a tiny elite. Thanks to Facebook, "the markets" (a.k.a. billionaire investors) even control human social interactions once considered sacrosanct from corporate intrusion. The Capitalist economy is all about the ascendancy of the 1% (and the 0.01% within that).
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/
I've just finished reading the claims and scanning the description and found no part of it that should be patentable; anybody reasonably skilled in statistics would end up with something similar or even identical. Heck, I'd even end up with something pretty much the same, and I have no formal background in statistics.
Also note that the patent names the required activities only (mostly it just lists potential sources of data); it does not explain the methods or mechanisms used to perform those activities.
Slashdot social media options: AIM, ICQ, Yahoo, Jabber and Mobile Text. Why no MySpace?
Statistics doesn't pigeon-hole you. It discovers what factors tend to influence people grouped with you, by how much, and how reliably. Like psychohistory, it only works on groups, the larger the better. The "pigeon-hole" is fuzzy and somewhat arbitrary. You still (maybe) have free will and are an individual... just like everybody else.
Pretty, pretty low.
This is patently absurd. In the UK, Equifax, Experian and Call Credit already sell income predictions based on statistical modelling of credit bureau information. How is switching the underlying data set in any way a unique or clever thing to do?
This is nothing more than a fancy regression algorithm.
It's a statistical measure. People who read such things generally have a low income. The fact that you specifically read them and don't have a low income is irrelevant; the advertisers don't care about you as an individual. The large number of people who do fit the profile make the advertising more lucrative to a degree which far overwhelms the small number of people like you who make it less lucrative.