<?xml version="1.0" standalone="yes"?>
<Paper uid="C00-2135">
  <Title>Acquisition of Phrase-level Bilingual Correspondence using Dependency Structure</Title>
  <Section position="2" start_page="0" end_page="0" type="intro">
    <SectionTitle>
1 Introduction
</SectionTitle>
    <Paragraph position="0"> Since the advent of statistical methods in Machine qh'anslation, the bilingual sentence alignmerit (Brown et al., 1991) or word alignment (Dagan et al., 1992) have been explored and achieved numerous success over the last decade.</Paragraph>
    <Paragraph position="1"> In coN;rasl,, fewer resull;s are reported in phrase-level correspondence. As word sequences are not translated literally a word for a word, acquiring phraseqevel correspondence still remains an important problem to be exploited.</Paragraph>
    <Paragraph position="2"> This paper proposes a method to extract phrase-level correspondence fi'om sentence-aligned parallel corpora using statistically probable dependency relations, i.e. head-modifier relations in a sentence.</Paragraph>
    <Paragraph position="3"> The distinct characteristics of our approach is two-fold. First, our approach uses dependency relations rather than alignment, cognate and/or position heuristics previously applied (Melamed, 1995). Our approach is based on the assumption that the word ordering and positions may not necessarily coincide between the two languages, but the dependency structure between words will be preserved. We believe that dependency relations offer richer linguistic clues (syntactic information) and are effcctive for language pairs with different word ordering constraints.</Paragraph>
    <Paragraph position="4"> Secondly, statistical dependency parsers are used to obtain candidate patterns. Previous methods mostly use rule-based parsers for preprocessing(Matsumoto et al., 1993), (Kitamura and Matsumoto, 1995). The progress in parsing technology are noteworthy, and in particular, various statistical dependency models have been proposed(Collins, 1997),, (Ratnaparkhi, 1997), (Charniak, 2000). It has an advantage over the rule-based counterpart in that it achieves wider coverage, does not need to care for consistency in rule writing, and is robust to domain changes.</Paragraph>
    <Paragraph position="5"> We conjecture that our approach improves coverage a.nd robustness by use of sl;atistical dependency parsers.</Paragraph>
    <Paragraph position="6"> In this paper, we aim to bwestigate tim efficacy of statistically probable dependency structure in finding phrase level bilingual correspondence. Though our discussion will proceed for English Japanese phrasal correspondence, the proposed approach is applicable to any pair of languages.</Paragraph>
    <Paragraph position="7"> This paper is organised as follows: In the next section, we present the overview of our approach. In Sections 3 and 4, components are elaborated in detail. In Section 5~ experiment and results are given. In Section 6, we compare our approach with related works, and finally our findings are concluded in Section 7.</Paragraph>
  </Section>
class="xml-element"></Paper>