SoylentNews is people
SoylentNews
from the I-wonder-if-they-track-sales-of-tin-foil? dept.
http://www.cio.com/article/2977027/intel-reveals-big-datas-dirty-little-secret.html
The article is entitled "Intel reveals big data's dirty little secret" but I read it a little bit differently.
From the article: "Companies are spending billions on tools and engineering to analyse big data, though many are hampered by one little problem: they still don't know what to do with all the data they collect."
This means that, of all the egregious breaches of personal privacy that companies regularly perform (the Target-knows-you're-pregnant-when-your-parents-don't story comes to mind), they have still only scratched the surface of making sense of your information, and using it effectively. Which means that, as Big Data gets people who actually know what they're doing, the more frightening the possibilities become, which is probably only a matter of time.
How would you feel about getting a bunch of targeted spam from divorce lawyers because your wife/husband's personal details were in the big Ashley Madison data leak, before you even heard about it? What if you were the guy who got drunk and put a profile up one time after a big fight but never followed up on it? This is why I don't have a Facebook account.
Original Submission
(Score: 5, Insightful) by looorg on Saturday August 29 2015, @04:22PM
"Companies are spending billions on tools and engineering to analyse big data, though many are hampered by one little problem: they still don't know what to do with all the data they collect"
What they don't know is that most of the data they keep gathering and storing is worthless junk, either they don't know or they know but they keep hoping for things to turn around. But no matter how hard they polish that giant data turd it's still going to be a turd. It's just not going to become golden just cause they find the right algorithm. There is no data-science-alchemy, not that it stops people from selling that -- the sales rep from IBM and others won't tell you that you have a data turd cause they want to keep selling your magic solutions. I have seen it in academic data, decades worth of data turned digital. I have seen it in private companies data. They gather so much data and most of it is just crap. They are holding onto it cause they one day believe that it will turn to data-gold. It's not, or I'll be utterly surprised if it did. But as long as data storage is cheap they'll keep doing it.
It's what the people that thought me refer to as a SISU (it looses a bit in translation but basically it means 'shit in, shit out') problem; bad data won't become good data just cause you massage it. It might look golden but it's still shit.
(Score: 0) by Anonymous Coward on Saturday August 29 2015, @05:10PM
I was going to post something similar to this but I see you beat me to it. I would add that this is a form of hoarding, much like dysfunctional people hoard. It does no good and causes companies to lose money by spending all their money on the hardware/software/people/electricity needed to keep these systems running.
(Score: 2) by Justin Case on Saturday August 29 2015, @05:26PM
> most of the data they keep gathering and storing is worthless junk... the sales rep from IBM and others won't tell you that you have a data turd cause they want to keep selling your magic solutions.
In my experience the biggest suckers for sales weasels are other sales weasels. By lying with every breath for years, their brains have rotted to the point where they no longer grasp the concept "truth". Therefore they are defenseless when someone says "buy my crap, it will help you sell your crap."
> SISU 'shit in, shit out'
Where I live it's GIGO: garbage in garbage out. Same idea. Anyway, you have a moral obligation to stuff the Big Data Hoarders full of as much garbage -- or shit -- as you can. And I hope they find it quite tasty, yes indeed!
(Score: 1) by dingus on Saturday August 29 2015, @09:19PM
This is true in some sense, but obviously the researchers at the NSA believe it's possible to turn it into a golden turd, so I'm inclined to follow their lead. Some of the top mathematicians and computer scientists in the world work there.
Of course, I could be wrong and they could be running their collection operations due to political reasons. In which case I guess the top mathematicians and computer scientists at the NSA are pretty pissed.
(Score: 1) by termigator on Saturday August 29 2015, @10:26PM
The data can be beneficial if you have specific queries in mind. I.e. You already have specific targets in mind and need to gather more information about those targets. Of course, for organizations like the NSA, such data gathering is unconstitutional, but it seems like the checks-and-balances of the U.S. govt has failed in this regard.
(Score: 2) by Justin Case on Sunday August 30 2015, @02:54AM
> The data can be beneficial if you have specific queries in mind. I.e. You already have specific targets^Wpresidential candidates in mind and need to gather more information about those targets^Wpresidential candidates.
FTFY
(Score: 2) by etherscythe on Sunday August 30 2015, @01:12PM
This may be true in some cases, but the Target story really brought home to me that as data collection techniques and partnerships improve, the risks are definitely real. In case you're not familiar with that story, I strongly recommend following up on it.
"Fake News: anything reported outside of my own personally chosen echo chamber"
(Score: 2) by looorg on Sunday August 30 2015, @02:40PM
I'm not saying there are not uses for big data, there are. But just as with any data collection you have to do a lot of work to make sure you are actually gathering needed data that you can somehow use for something interesting. A lot of these big data projects seem to start in the wrong end of things. Normally one would start with the idea or the question and then gather data, they start with gathering (or a giant pile of previously collected data) and then they wonder if they can do something interesting with what they got. Sometime they can, no doubt about that - but they also will end with a lot of shit data that is just wasting space.
I'm not overly familiar with the Target case but I a quick scan seems to indicate that it's millions of records of sales, CC numbers and personal information from the customers etc. No doubt interesting things could be done with that. You can always find interesting cases, I was making more of a general statement.