<?xml version="1.0" standalone="yes"?>
<Paper uid="P04-3015">
  <Title>Hierarchy Extraction based on Inclusion of Appearance</Title>
  <Section position="9" start_page="0" end_page="0" type="concl">
    <SectionTitle>
9 Conclusion
</SectionTitle>
    <Paragraph position="0"> We have proposed a method of automatically extracting hierarchies based on an inclusion relation of appearance patterns from corpora. In this paper, we attempted to extract objective hierarchies of abstract nouns co-occurring with adjectives in Japanese. In our experiment, we showed that complementary similarity measure can extract a kind of hierarchy from corpora, though it is a similarity measure developed for the recognition of degraded machine-printed text. Also, we can find interesting hierarchies which suit human intuition, though they are different from exact hierarchies. Kanzaki et al. (2004) have applied our approach to verify classification of abstract nouns by using self-organization map. We can look a suitability of our result at that work.</Paragraph>
    <Paragraph position="1"> In our future work, we will use our approach for other parts of speech and other types of word.</Paragraph>
    <Paragraph position="2"> Moreover, we will compare with current alternative approaches such as those based on sentence patterns.</Paragraph>
  </Section>
class="xml-element"></Paper>