<?xml version="1.0" standalone="yes"?>
<Paper uid="W06-1621">
  <Title>Lexical Reference: a Semantic Matching Subtask</Title>
  <Section position="11" start_page="177" end_page="177" type="concl">
    <SectionTitle>
1 QNX Software Systems Ltd., a leading provider of real-time software and ser-
</SectionTitle>
    <Paragraph position="0"> vices to the embedded computing market, is pleased to announce the appointment of Mr. Sachin Lawande to the position of vice president, engineering services. null  is better at identifying phrase and context examples. The combined WordNet and Bayesian models' strength can be explained by the quite different behaviors of the two models - the Word-Net model seems to be better in identifying the word-to-word explicit reference examples while the Bayesian model is better in modeling the contextual implied references.</Paragraph>
  </Section>
class="xml-element"></Paper>