<?xml version="1.0" standalone="yes"?>
<Paper uid="E99-1025">
  <Title>New Models for Improving Supertag Disambiguation</Title>
  <Section position="9" start_page="193" end_page="193" type="concl">
    <SectionTitle>
6 Conclusions
</SectionTitle>
    <Paragraph position="0"> We have introduced two different kinds of models for the task of supertagging. Contextual models show that features for accurate supertagging only produce improvements when they are appropriately combined. Among these models were: a one pass head model that reduces propagation of head detection errors of previous models by using supertags themselves to identify heads; a mixed model that combines use of local and long distance information; and a classifier combination model that ameliorates the sparse data problem that is worsened by the introduction of many new features. These models achieve better supertagging accuracies than previously obtained. We have also introduced class based models which trade a slight increase in ambiguity for significantly higher accuracy. Different class based methods are discussed, and the tradeoff between accuracy and ambiguity is demonstrated.</Paragraph>
    <Paragraph position="1"> 7Again, for the class C assign to a given word w~, we consider only those tags ti E C for which/5(wdti) &gt; 0.</Paragraph>
  </Section>
class="xml-element"></Paper>