<?xml version="1.0" standalone="yes"?>
<Paper uid="P05-1016">
  <Title>Inducing Ontological Co-occurrence Vectors</Title>
  <Section position="8" start_page="131" end_page="131" type="concl">
    <SectionTitle>
7 Conclusions
</SectionTitle>
    <Paragraph position="0"> We presented a framework for inducing ontological feature vectors from lexical co-occurrence vectors. Our method does not require the disambiguation of text. Instead, it relies on the principle of distributional similarity and the fact that polysemous words that are similar in one sense tend to be dissimilar in their other senses. On the task of attaching new words to WordNet using our framework, our experiments showed that the first attachment has 73.9% accuracy and that a correct attachment is in the top-5 attachments with 81.3% accuracy.</Paragraph>
    <Paragraph position="1"> We believe this work to be useful for a variety of applications. Not only can sense selection tasks such as word sense disambiguation, parsing, and semantic analysis benefit from our framework, but more inference-oriented tasks such as question answering and text summarization as well.</Paragraph>
    <Paragraph position="2"> We hope that this work will assist with the development of other large-scale and internally consistent collections of semantic information.</Paragraph>
  </Section>
class="xml-element"></Paper>