<?xml version="1.0" standalone="yes"?>
<Paper uid="W06-3105">
  <Title>Why Generative Phrase Models Underperform Surface Heuristics</Title>
  <Section position="5" start_page="37" end_page="37" type="concl">
    <SectionTitle>
4 Conclusion
</SectionTitle>
    <Paragraph position="0"> Re-estimating phrase translation probabilities using a generative model holds the promise of improving upon heuristic techniques. However, the combinatorial properties of a phrase-based generative model have unfortunate side effects. In cases of true ambiguity in the language pair to be translated, parameter estimates that explain the ambiguity using segmentation variables can in some cases yield higher data likelihoods by determinizing phrase translation estimates. However, this behavior in turn leads to errors at decoding time.</Paragraph>
    <Paragraph position="1"> We have also shown that some modest benefit can be obtained from re-estimation through the blunt instrument of interpolation. A remaining challenge is to design more appropriate statistical models which tie segmentations together unless sufficient evidence of true non-compositionality is present; perhaps such models could properly combine the benefits of both current approaches.</Paragraph>
  </Section>
class="xml-element"></Paper>