GeoQuery is a dataset used for benchmarking semantic parsers. It contains 880 queries about USA geography. The queries are in Prolog format, for example:
answer(A,longest(A,(river(A),traverse(A,B),const(B,countryid(usa)))))
(this represents a question for the longest river that passes through the us).
I have read several papers about the GeoQuery, but they don't seem to give a formal definition of the grammar underlying the dataset. The email address in the dataset page seem to be dysfunctional. So, my question (addressed primarily to researchers who have worked with the GeoQuery dataset before) is twofold:
- Is there a formal definition of the GeoQuery language?
- Is this language context-free?
EDIT:
It seems that, if the number of variables (A, B, C...) is unlimited, then GeoQuery is not context-free, because a CFG cannot make sure that two variables in different places are identical (this is similar to the {ww} language, which is not context-free).
But what if we limit ourselves to A, B and C?