>>5372652 (pb)
It's good, but still very manual. The maps of qmap.pub produce some interesting results but I don't find them totally coherent.
What is really needed I think is some kind of learning system that can collect (public domain) events on a large scale and then correlate them, flagging events that might otherwise seem unrelated, such as tweets, FB posts and news articles.
(pic related to use with your libtard alter-ego)