1. 5

An interesting study showing the breath of information that can be inferred from mobile communications meta data.

  1.  

  2. 4

    The “ethical considerations” section is the right way to do this kind of study. Infinitely more responsible than the recent OK Cupid thing.

    This includes a very interesting, obvious-in-retrospect finding about customer service lines drastically increasing the connectivity of the graph. That’s relevant because hop count is one of the purported protections in “real” surveillance.

    The rest of this is a good read as well, and I do recommend it to anyone interested in this sort of analysis. It’s also very important to have public research informing this policy conversation.