Several researchers have tried to evaluate the tree induction method itself, to precisely answer questions such as is it possible to build optimal trees?, how good are particular heuristics (feature evaluation rules or pruning methods)? Most of these investigations are theoretical, though there have been a few recent empirical ones.