Abstract

Prepositional phrase attachment has long been considered one of the most difficult tasks in automated syntactic parsing of natural language text. In this thesis, we examine several aspects of what has become the dominant view of PP attachment in natural language processing with an eye toward extending this view to a more realistic account of the problem. In particular, we take issue with the manner in which most PP attachment work is evaluated, and the degree to which traditional assumptions and simplifications no longer allow for realistically meaningful assessments. We also argue for looking beyond the canonical subset of attachment problems, where almost all attention has been focused, toward a fuller view of the task, both in terms of the types of ambiguities addressed and the contextual information considered.