Bizarre things big data tells you you really really want

Page may contain affiliate links. Please see terms for details.

ozboz

Guru
CED92C54-556D-4C10-99D0-4B1EAC17123B.jpeg


So oddly enough this ad appeared just now as I opened this thread ,
Do I need to elimate data silos in Life Sciences ?
Tbh , I didn’t know I had data silos ,
 

Ming the Merciless

There is no mercy
No, don't like cats.
I know how ebay made the link but it shows how wrong and bonkers algorithms etc can be.

Well it's not necessarily wrong and if it based on big data then it won't be based purely on your data. If it was based on your data then it'd be called tiny data. Let's say that big data says that if we advertise this product to people "like yourself" (determined via algorithm chrunching lots of data about lots of peoples buying habits and other stuff) then 0.5% will buy the hammock. Then lets say a random advert has a 0.02% chance of making you buy said product. Then it is worth while them serving that advert rather than something else. Analysing lots of data to come to conclusions doesn't mean 100% hit rate it just means a significant improvement in hit rate (over random stuff) to improve their profits.
 
Last edited:
OP
OP
Blue Hills
Interesting valid points yukon but the mail said it was based on MY interests and previous searches.

And a cat, let alone any treats for said spoiled brats, remains a 100 percent non purchase for me.
 
OP
OP
Blue Hills
OK yukon.

It's little data used by big interests.

Which makes it even more bizarre.

May get round to changing the thread title.
 

Ming the Merciless

There is no mercy
More importantly I'm worried about the Internet Kittens not getting their hammocks...
 
...big data says that if we advertise this product to people "like yourself" (determined via algorithm chrunching lots of data about lots of peoples buying habits and other stuff)...

Well, people like me go to "live, work and study as an undergraduate at Yale".

So there.

As an aside, I didn't know locksmiths had to have a degree...
 
Top Bottom