Using decision trees to select the grammatical relation of a noun phrase

We present a machine-learning approach to modeling the distribution of noun phrases (NPs) within clauses with respect to a fine-grained taxonomy of grammatical relations. We demonstrate that a cluster of superficial linguistic features can function as a proxy for more abstract discourse features that are not observable using state-of-the-art natural language processing. The models constructed for actual texts can be used to select among alternative linguistic expressions of the same propositional content when generating discourse.