Not exactly the same thing, but several tools in the Government rely on "crowd-sourcing" entity extraction and disambiguation vs. using an automated approach -- Axis Pro, Palantir, etc.<p>You get high quality named entity tagging (since humans are doing it), but it takes a lot longer than doing it by machine, and it has a couple of downsides -- like scaling poorly, involving lots of staff doing entity tagging instead of real work, a noticeable delay between incoming reports and being able to use them (since they aren't tagged and those systems pretty much require tagged documents to do anything useful with them), as well as people fighting over the disambiguation of certain entities.