<?xml version="1.0" standalone="yes"?>
<Paper uid="E06-1042">
  <Title>A Clustering Approach for the Nearly Unsupervised Recognition of Nonliteral Language[?]</Title>
  <Section position="7" start_page="335" end_page="335" type="concl">
    <SectionTitle>
6 Conclusion
</SectionTitle>
    <Paragraph position="0"> In this paper we presented TroFi, a system for separating literal and nonliteral usages of verbs through statistical word-sense disambiguation and clustering techniques. We suggest that TroFi is applicable to all sorts of nonliteral language, and that, although it is currently focused on English verbs, it could be adapted to other parts of speech and other languages.</Paragraph>
    <Paragraph position="1"> We adapted an existing word-sense disambiguation algorithm to literal/nonliteral clustering through the redefinition of literal and nonliteral as word senses, the alteration of the similarity scores used, and the addition of learners and voting, SuperTags, and additional context.</Paragraph>
    <Paragraph position="2"> For all our models and algorithms, we carried out detailed experiments on hand-annotated data, both to fully evaluate the system and to arrive at an optimal configuration. Through our enhancements we were able to produce results that are, on average, 16.9% higher than the core algorithm and 24.4% higher than the baseline.</Paragraph>
    <Paragraph position="3"> Finally, we used our optimal configuration of TroFi, together with active learning and iterative augmentation, to build the TroFi Example Base, a publicly available, expandable resource of literal/nonliteral usage clusters that we hope will be useful not only for future research in the field of nonliteral language processing, but also as training data for other statistical NLP tasks.</Paragraph>
  </Section>
class="xml-element"></Paper>