I find myself with a number of colleagues looking at ecological monitoring, so this article about using web crawling for ecological monitoring caught my eye. You might remember the trends tool at Google that got a lot of attention last fall that mined the queries people typed into Google, correlated them with known cases of the flu, and then watched new queries as they came in to try to spot new locations where the flu had cropped up as it was starting.
The idea here is the same – take the data that has already been maintained by Google or other search engines to find patterns in queries that correlate with ecological decline. Then, monitor ongoing queries to get early alerts of problems when it may still be early enough to address the situation.
A very cool idea if it could work. But it is also worth asking what other problems you might want to detect in this way. Criminal or terrorist activity are obvious targets. If that seems abstract and distant from you, what about schools mining the queries of their students for evidence that they are likely plagiarists and ought to have their work more closely scrutinized. It is possible to bring these technologies quite close to home.