Seems like a subset of CYC - attempting to gather causal data rather than declarative data in general.

It's a bit odd that their paper doesn't even mention CYC once.