[Sociam-soton] Re: how twitter knows you're depressed : Eric Horvitz at MSR

[Leslie Carr]

The URL for the paper (which didn't seem to be in the time report on my phone) is at http://research.microsoft.com/en-us/um/people/horvitz/depression_populations_websci_2013.pdf

I wonder what the health scientists make of the study? Is it cheap twitterology or something more profound? 

Note that the study starts by identifying depression sufferers using mechanical Turk - bootstrapping the development of one putative social machine with another!

Sent from my iPhone

> On 28 Jan 2014, at 00:35, "electronic Max" <emax at ecs.soton.ac.uk> wrote:
> 
> MSR Guru Eric Horvitz (also Nigel's friend) reported new results where
> they did a careful study of the Twitter posts made over time by people
> who were (later) diagnosed with depression. And it made Time magazine,
> among other things >
> http://business.time.com/2014/01/27/how-twitter-knows-when-youre-depressed/
> 
> Looks like a well constructed study and it will be good to read the
> paper - also very interesting for both PDS and Web Observatory
> applications.
> 
> Best,
> Max
> _______________________________________________
> Sociam-soton mailing list
> Sociam-soton at ecs.soton.ac.uk
> http://mailman.ecs.soton.ac.uk/mailman/listinfo/sociam-soton