<?xml version="1.0" standalone="yes"?>
<Paper uid="W05-0831">
  <Title>Novel Reordering Approaches in Phrase-Based Statistical Machine Translation</Title>
  <Section position="9" start_page="173" end_page="173" type="concl">
    <SectionTitle>
6 Conclusion
</SectionTitle>
    <Paragraph position="0"> In this paper, we described a reordering framework which performs source sentence reordering on word level. We suggested to use optimal alignment functions for monotonization and improvement of translation model training. This allowed us to translate monotonically taking a reordering graph as input.</Paragraph>
    <Paragraph position="1"> We then described known and novel reordering constraints and their efficient finite-state implementations in which the reordering graph is computed ondemand. We also utilized weighted permutations.</Paragraph>
    <Paragraph position="2"> We showed that our monotonic phrase-based translation approach effectively makes use of the reordering framework to produce quality translations even from languages with significantly different word order. On the Japanese-to-English and Chinese-to-English IWSLT tasks, our system performed at least as well as the best machine translation system.</Paragraph>
  </Section>
class="xml-element"></Paper>