<?xml version="1.0" standalone="yes"?>
<Paper uid="W04-0307">
  <Title>A Statistical Constraint Dependency Grammar (CDG) Parser</Title>
  <Section position="3" start_page="0" end_page="0" type="intro">
    <SectionTitle>
1 Introduction
</SectionTitle>
    <Paragraph position="0"> Statistical parsing has been an important focus of recent research (Magerman, 1995; Eisner, 1996; Charniak, 1997; Collins, 1999; Ratnaparkhi, 1999; Charniak, 2000). Several of these parsers generate constituents by conditioning probabilities on non-terminal labels, part-of-speech (POS) tags, and some headword information (Collins, 1999; Ratnaparkhi, 1999; Charniak, 2000). They utilize non-terminals that go beyond the level of a single word and do not explicitly use lexical features. Collins' Model 2 parser (1999) learns the distinction between complements and adjuncts by using heuristics during training, distinguishes complement and adjunct non-terminals, and includes a probabilistic choice of left and right subcategorization frames, while his Model 3 parser uses gap features to model wh-movement. Charniak (Charniak, 2000) developed a state-of-the-art statistical CFG parser and then built an effective language model based on it (Charniak, 2001). But his parser and language model were originally designed to analyze complete sentences. Among the statistical dependency grammar parsers, Eisner's (1996) best probabilistic dependency model used unlabeled links between words and their heads, as well as between words and their complements and adjuncts. However, the parser does not distinguish between complements and adjuncts or model whmovement. Collins' bilexical dependency grammar parser (1999) used head-modifier relations between pairs of words much as in a dependency grammar, but they are limited to relationships between words in reduced sentences with base NPs.</Paragraph>
    <Paragraph position="1"> Our research interest focuses on building a high quality statistical parser for language modeling. We chose CDG as the underlying grammar for several reasons. Since CDGs can be lexicalized at the wordlevel, a CDG parser-based language model is an important alternative to CFG parser-based models, which must model both non-terminals and terminals. Furthermore, the lexicalization of CDG parse rules is able to include not only lexical category information, but also a rich set of lexical features to model subcategorization and wh-movement. By using CDG, our statistical model is able to distinguish between adjuncts and complements. Additionally, CDG is more powerful than CFG and is able to model languages with crossing dependencies and free word ordering.</Paragraph>
    <Paragraph position="2"> In this paper, we describe and evaluate a statistical CDG parser for which the probabilities of parse prefix hypotheses are incrementally updated when the next input word is available, i.e., it parses incrementally. Section 2 describes how CDG represents a sentence's parse and then defines a Super-ARV, which is a lexicalization of CDG parse rules used in our parsing model. Section 3 presents the parsing model, while Section 4 motivates the evaluation metric used to evaluate our parser. Section 5 presents and discusses the experimental results.</Paragraph>
  </Section>
class="xml-element"></Paper>