<?xml version="1.0" standalone="yes"?>
<Paper uid="W06-3119">
  <Title>Syntax Augmented Machine Translation via Chart Parsing</Title>
  <Section position="7" start_page="140" end_page="140" type="concl">
    <SectionTitle>
6 Conclusions
</SectionTitle>
    <Paragraph position="0"> In this work we applied syntax based resources (the target language parser) to annotate and generalize phrase translation tables extracted via existing phrase extraction techniques. Our work reaffirms the feasibility of parsing approaches to machine translation in a large data setting, and illustrates the impact of adding syntactic categories to drive and constrain the structured search space.</Paragraph>
    <Paragraph position="1"> While no improvements were available at submission time, our subsequent performance highlights the importance of tight integration of n-gram language modeling within the syntax driven parsing environment. Our translation system is available open-source under the GNU General Public License at: www.cs.cmu.edu/~zollmann/samt</Paragraph>
  </Section>
class="xml-element"></Paper>