In previous posts we discussed many aspects of differential privacy: what it is, what it is useful for, and how it is applied to data analysis problems. All of those ideas can be applied once you get your hands on a whole dataset. What if the data you are interested in extracting insights from belongs to mutually distrusting organizations? For example, say you run a pumpkin spice latte stand and are wondering if your pumpkin spice supplier is overcharging you compared to the industry-wide average. You are willing to participate in a study that computes this average, but not comfortable giving
Source: nist.gov