Prepositional Phrase Attachment in Shallow Parsing

International Conference RANLP 2009 - Borovets, Bulgaria, pages 12–17 Prepositional Phrase Attachment in Shallow Parsing Vincent Van Asch CLiPS - Computational Linguistics University of Antwerp Prinsstraat 13 B-2000 Antwerpen, Belgium Vincent.VanAsch@ua.ac.be Walter Daelemans CLiPS - Computational Linguistics University of Antwerp Prinsstraat 13 B-2000 Antwerpen, Belgium Walter.Daelemans@ua.ac.be Abstract In this paper we extend a shallow parser [6] with prepositional phrase attachment. Although the PP attachment task is a well-studied task in a discriminative learning context, it is mostly ad- dressed in the context of artificial situations like the quadruple classification task [18] in which only two possible attachment sites, each time a noun or a verb, are possible. In this pa- per we provide a method to evaluate the task in a more natural situation, making it possible to compare the approach to full statistical pars- ing approaches. First, we show how to extract anchor-pp pairs from parse trees in the GENIA and WSJ treebanks. Next, we discuss the exten- sion of the shallow parser with a PP-attacher. We compare the PP attachment module with a statistical full parsing approach [4] and ana- lyze the results. More specifically, we investi- gate the domain adaptation properties of both approaches (in this case domain shifts between journalistic and medical language). Keywords prepositional phrase attachment, shallow parsing, machine learning of language 1 Introduction Shallow parsing (also called partial parsing) is an ap- proach to language processing that computes a basic analysis of sentence structure rather than attempting full syntactic analysis. Originally defined by Abney [1] as a task to be solved with handcrafted regular expressions (finite state methods) and limited to finding basic (non- recursive) phrases in text, the label shallow parsing has meanwhile broadened its scope to machine learn- ing methods and to a set of related tasks including part of speech tagging, finding phrases (chunking), clause identification, grammatical role labeling, etc. Especially the machine learning approach to shallow parsing, pioneered by Ramshaw and Marcus [17] has been investigated intensively, in part because of the availability of benchmark datasets and competitions (CoNLL shared tasks 1999 to 2001)1. 1 See http://ifarm.nl/signll/conll/ It has been argued in [10] and by others that full parsing often provides too much (or not enough) infor- mation for some frequent natural language processing tasks. For example, for information retrieval, find- ing basic NPs and VPs is arguably sufficient, and for information extraction and other text mining tasks, finding syntactic-semantic relations between verbs and base NPs (who did what when and where) is more im- portant than having an elaborate configurational syn- tactic analysis, provided this shallow analysis can be computed in a deterministic, efficient, robust, and ac- curate way. Another advantage is that the modules in a machine learning based shallow parser can be trained independently, and allow the inclusion of more information sources (input features) than is possible in statistical parsing (because of sparse data problems). This flexibility in feature engineering, inherent in dis- criminative, supervised learning approaches to shallow parsing should make the approach more flexible, e.g. …

Content truncated to 500 words