I think we received a total of 14 different National Science Foundation grants, at least, to support various aspects. The very first funding we received for eBird was a National Science Foundation grant for informal science education. It began as an education-focused initiative.
It has transformed. Now, most of the funding we get is in the area of machine learning and statistics, and the interface of those two. Part of the reason for that is that the fundamental challenges in analyzing epidemiological data are very similar to the ones we're trying to work out in citizen science data. I spoke about this a bit before. You basically have a sensor network, where you may have doctors or birdwatchers who are reporting in slightly different rates. You're trying to understand the bias in both of those—both from your sensor and then what the truth is.
Machine learning has a lot of very good applications in terms of neural network models to not necessarily understand the drivers of those, where statistical frameworks are—