Beautiful Data

Relevance: What Matters and to Whom?

Now we have lots of data coming at us. Take a moment to assess all the sights and sounds in your immediate environment: the background hum of the computer, the music playing, the cartoons on the TV, and the kids trying to steal one another's toys. As humans, we have internal relevance-detection algorithms that assess all of the presented data, and alert us when some observations cause us to take notice and possibly react. Sort of like ignoring the kids and their ongoing bickering until it threatens to escalate into violence. It's that one move, by one kid, that alerts us that it is time to intercede.

How one exploits technology to calculate relevance and when to allow it to automatically register an alarm is crucial. On the technology side, the objective is a state in which the very next item in an alarm queue is the next most important item for review. There is also no reason to produce more alarms than there are available resources (for example, analysts, systems) to deal with them. Risk-assessment engines, for example, must be configured to produce alarms appropriate to one's individualized risk, staffing, and ability to respond. If resources increase, one can increase alarm sensitivity.