<?xml version="1.0" standalone="yes"?>
<Paper uid="W04-1808">
  <Title>Discovering Synonyms and Other Related Words</Title>
  <Section position="8" start_page="0" end_page="0" type="concl">
    <SectionTitle>
7 Conclusion
</SectionTitle>
    <Paragraph position="0"> We have demonstrated that it is feasible to calculate similarities between words using a full dependency syntactic feature space. We have also introduced similarity recalculation during clustering as a fast approximation of the high-dimensional feature space. In addition we introduced a way to use translation dictionaries for evaluating the rate of synonymy found in the word clusters, which is useful for languages that do not yet have publicly available thesaurus resources like WordNet. Finally we have shown that 69-79% of the words in the discovered clusters are useful for thesaurus construction.</Paragraph>
  </Section>
class="xml-element"></Paper>