<?xml version="1.0" standalone="yes"?>
<Paper uid="W04-0815">
  <Title>Dependency Based Logical Form Transformations</Title>
  <Section position="7" start_page="0" end_page="0" type="concl">
    <SectionTitle>
6 Conclusions
</SectionTitle>
    <Paragraph position="0"> Dependency grammars provide a natural and intuitive solution to the task of logical form identific ation. We have managed to demonstrate relatively good overall performance on the given task with minimal additional processing and a very small amount of training data.</Paragraph>
    <Paragraph position="1"> It is argued that a dependency grammar based parse provides a rich source of knowledge that is suitable for the transformation of English sentences into a logical form. It would appear that there is to a large extent enough information embedded within the parser's output to achieve the desired outcome. It is however apparent that other types of information could further improve the solution. These types of information include named entity recognition and multi-word phrase detection.</Paragraph>
  </Section>
class="xml-element"></Paper>