Prepositional Phrase Attachment through a Backed-off Model

Recent work has considered corpus-based or statistical approaches to the problem of prepositional phrase a t tachment ambiguity. Typically, ambiguous verb phrases of the form v rip1 p rip2 are resolved through a model which considers values of the four head words (v, n l , p and 77,2). This paper shows that the problem is analogous to n-gram language models in speech recognition, and that one of the most common methods for language modeling, the backed-off estimate, is applicable. Results on Wall Street Journal data of 84.5% accuracy are obtained using this method. A surprising result is the importance of low-count events ignoring events which occur less than 5 times in training data reduces performance to 81.6%. 1 I n t r o d u c t i o n Prepositional phrase a t tachment is a common cause of structural ambiguity in natural language. For example take the following sentence: Pierre Vinken, 61 years old, jo ined the board as a nonexecutive director. The PP 'as a nonexecutive director' can either at tach to the NP ' the board' or to the VP 'joined', giving two alternative structures. (In this case the VP at tachment is correct): NP-attach: (joined ((the board) (as a nonexecutive director))) VP-attach: ((joined (the board)) (as a nonexecutive director)) Work by Ratnaparkhi , Reynar and Roukos [RRR94] and Brill and Resnik [BR94] has considered corpus-based approaches to this problem, using a set of examples to train a model which is then used to make a t tachment decisions on test data. Both papers describe methods which look at the four head words involved in the a t tachment the VP head, the first NP head, the preposition and the second NP head (in this case joined, board, as and director respectively). This paper proposes a new statistical method for PP-a t tachment disambiguation based on the four head words.