<?xml version="1.0" standalone="yes"?>
<Paper uid="C00-2092">
  <Title>Data-Oriented Translation</Title>
  <Section position="9" start_page="639" end_page="639" type="concl">
    <SectionTitle>
6 Conclusion
</SectionTitle>
    <Paragraph position="0"> In this article, we have presented a new approach to machine translation: the Data-Oriented Translation model. This method uses linked subtree pairs for creating a derivation of a sentence. Each subtreepair consists of two trees: one in the source language and one in the target language. Using these subtree pairs, we can form a derivation of a given source sentence, which can then be used to form a target sentence. The probability of a translation can  then be calculated as the total probability of all the derivations that form tiffs translation.</Paragraph>
    <Paragraph position="1"> The computational aspects of DOT have been discussed, where we introduced a way to reform each subtree pair into a productive rewrite role so that well-known parsing algorithms can be used. We del:ermine the best translation by Monte Carlo sampling. null We have discussed the results of some pilot experiments with a part of the Verbmobil corpus, and showed a method of evaluating them. The ewfluation showed that DOT produces less correct translation than Systran, but also less incorrect translations. We expected to see an increase in performance as we increased the depth of subtree pairs used, but this was not the case.</Paragraph>
    <Paragraph position="2"> Finally, we supplied some topics which art open l'or future research.</Paragraph>
  </Section>
class="xml-element"></Paper>