Add comments here for "Feature extraction languages for propositionalized relational learning" by Chad Cumby and Dan Roth.
From James Cussens: There are some issues early on in your paper which are preventing me from continuing on with the rest of it. Probably I've misread or misunderstood something. Help me!
Section 4.1 According to your Definition 4.1, if x is a variable, then p(x) is an atomic formula and hence also a formula. But at the end of Section 4.1 you assert that all formulae are closed. p(x) clearly isn't. So what's going on?
Section 4.2 You state that D is a collection of elements, and that constants in $\cal R$ denote elements of D. This I understand. But then you go on to say that D can be thought of as the collection of all ground atoms in $\cal R$. I don't understand how something can be denoted by a constant and a formula at the same time.