<?xml version="1.0" standalone="yes"?>
<Paper uid="C94-1049">
  <Title>CO-OCCURRENCE VECTORS FROM CORPORA VS. DISTANCE VECTORS FROM DICTIONARIES</Title>
  <Section position="7" start_page="307" end_page="308" type="concl">
    <SectionTitle>
5 Conclusion
</SectionTitle>
    <Paragraph position="0"> * A comparison was nlade of co-occnrrence vectors from large text corpora and of distance vectors from dictionary delinitions.</Paragraph>
    <Paragraph position="1"> * For tile word sense disambiguation based on the context simihtrity, co-occurrence vectors fl'om tile 1987 Wall Street Journal (20M total words) was advantageous over distance vectors from the Collins l,;nglish Dictionary (60K head words + 1.6M definition words).</Paragraph>
    <Paragraph position="2"> * For learning positive or negalive meanings from example words, distance vectors gave remarkably higher precision than co-occurrence vectors. This suggests, though further investigation is required, that distance w:ctors contain some different semantic information from co-occurrence vectors.</Paragraph>
    <Paragraph position="3">  I)ependence on vector dimension for (i) co-occurrence veetors and (ii) distance vectors. context size: 10, examples: 10/sense, corpus size for co-oe, vectors: 20M word.</Paragraph>
  </Section>
class="xml-element"></Paper>