Story

The Fallacies of Big Data

Preview

Good article, but a bit biased on its own (Score: 1)

by quadrox@pipedot.org on 2014-03-31 08:32 (#W5)

I fully agree with the overal message of the article, pointing out that just having a lot of data does not automatically prevent sampling errors/sampling bias and/or other fallacies.

That being said, I think they were downplaying what Target seems to have achieved a bit too much. Of course the system is bound to produce some false positives, but given the criteria described it does seem reasonable that they can make a quite good assessment of pregnancy. Granted, without having access to Targets systems we cannot know for sure how well it works, but the article seems to strongly imply that it doesn't work, and indeed cannot work.

Re: Good article, but a bit biased on its own (Score: 2, Interesting)

by zafiro17@pipedot.org on 2014-03-31 10:18 (#W8)

Good point about how the Target article fails to describe how many misses there are. I always get a good laugh out of the ads Google's fancy algorithms think I should see. I am regularly bombarded with ads for things I've already bought, for example.

There's an old quote by Pico Ayer (I can't find it at the moment but probably discovered it through that old website from the late 90s, the Utne Cafe), in which he worries that people are confusing information for knowledge, and knowledge for wisdom, and that although the modern technologies provide us tons of information they don't provide much knowledge, and far less wisdom. I'd kick that one level further to say that before even information, they drown us in data, which is totally worthless unless you know what you're looking for, or are trained in noticing things you were not suspecting.

Moderation

Time Reason Points Voter
2014-04-11 11:57 Interesting +1 ploling@pipedot.org

Junk Status

Not marked as junk