<?xml version="1.0" standalone="yes"?>
<Paper uid="W00-0714">
  <Title>Using Perfect Sampling in Parameter Estimation of a Whole Sentence Maximum Entropy Language Model*</Title>
  <Section position="6" start_page="80" end_page="81" type="concl">
    <SectionTitle>
5 Conclusion and future works
</SectionTitle>
    <Paragraph position="0"> We have presented a different approach to the sampling step needed in the parameter estimation of a WSME model. Using this technique, we have obtained a reduction of 30% in the perplexity of the WSME model over the base-line  trigram model and an improvement of 2% over the model trained with MCMC techniques. We are extending our experiments to a major corpus: the Wall Street Journal corpus and using a set of features which is more general, including features that reflect the global structure of the sentence.</Paragraph>
    <Paragraph position="1"> We are working on introducing the grammatical information contained into the sentence to the model; we believe that such information improves the quality of the model significantly.</Paragraph>
  </Section>
class="xml-element"></Paper>