A quick note on yet another mobile sensor for social psychology. Obviously, this is a very important topic for both social research and personal technology.
In an earlier post I discussed Alex Pentland’s work, which he recounts in his recent book Social Physics.
It is not surprising that there is more than one fish in this sea.
A second effort, apparently just as well-developed and similarly open is “Emotion Sense” http://emotionsense.org/ (out of Sheffield).
I haven’t tried or examined the software in detail, but it looks very similar to the toolkits from MIT. They have several papers, including the original report from 2010. They now have a company that is doing things with the technology, so it isn’t possible to know exactly what they are up to.
Several points come out of looking at their software.
First, the general design is for you to upload your data to their server/cloud, and they intend to use it to refine their models. I.e., the Android app is “free”, except you let them have all your data. Hmm.
It looks like I could, in principle, make my own server to collect the data, but they haven’t published any of the analytics. So no one can replicate their results independently, nor, by the way, get much value out of the “free” software they give away. Sigh.
A second lead-footed design feature is to require the users to connect via facebook. This gives them a ton of useful info on you, but kind of locks out the won’t-use-facebook crowd, such as me. So I can’t use the app, because I never will agree to facebook’s terms.
Third, the software operates rather opaquely, collecting data continuously without any obvious indication to the user. At least, that is what they say it does, and it makes sense. Notably, it is listening with the microphone, and gathering location and movement data. It might use the camera, too, I don’t know.
Their models map this information to self reports and other info, to infer mood, well being, etc. This is not a bug, this is exactly what the app is intended to do.
More than one commentator has noted that this is creepy. In particular, I’m creeped out by the opacity. I really don’t like my phone spying on me in ways I don’t understand, no matter why.
It is not so much what this group is trying to do—they only intend to feed the inferences about you back to you—the big issue is that if these guys can do it then you know ad agencies and Big Brother in general can certain do the same thing.
I will return to this topic in the future, with some longer book reviews.
Robert McGrath's Blog 