What is the status of this effort now?


#1

This dataset clearly took a lot of time and effort (and expense) to build. But we have always found a lot of anomalies in it, and whereas the currently publicly available CAMEO-based automated datasets clearly differ, we find that (for instance) the two GDELT datasets and ICEWS tend to reveal similar signals, whereas Terrier does not. For an example - see

, which shows factual negative great power assertiveness across these great powers.
I should also add that we have also found that performance of some our models (in which we use both structural variables AND event datasets) improves when we add GDELT, for instance. But we haven’t even tried with Terrier, because of the various very strange findings it gives us.
Before we fully abandon Terrier, we’d still love to find out what its actual status now is. Are you still working on it (you mention somewhere this was a beta version). Do you still have funding? Will there ever be a realtime version?
We’d be grateful if you could clear this up. Thanks much