<?xml version="1.0" standalone="yes"?>
<Paper uid="P06-2030">
  <Title>Using Bilingual Comparable Corpora and Semi-supervised Clustering for Topic Tracking</Title>
  <Section position="9" start_page="237" end_page="237" type="concl">
    <SectionTitle>
5 Conclusion
</SectionTitle>
    <Paragraph position="0"> In this paper, we addressed the issue of the difference in sizes between positive and negative training stories for the tracking task, and investigated the use of bilingual comparable corpora and semi-supervised clustering. The empirical results were encouraging. Future work includes (i) extending the method to an incremental approach for extracting story pairs, (ii) comparing our clustering method with the other existing methods such as X-means(Pelleg, 2000), and (iii) applying the method to the TDT4 for quantitative evaluation.</Paragraph>
  </Section>
class="xml-element"></Paper>