<?xml version="1.0" standalone="yes"?>
<Paper uid="W06-2403">
  <Title>Automatic Extraction of Chinese Multiword Expressions with a Statistical Tool</Title>
  <Section position="6" start_page="22" end_page="22" type="concl">
    <SectionTitle>
5 Conclusion
</SectionTitle>
    <Paragraph position="0"> In this paper, we have reported on our experiment of automatic extraction of Chinese MWEs using a statistical tool originally developed for English. Our statistical tool produced encouraging results, although further improvement is needed to become practically applicable for MT system in terms of recall. Indeed, for some constrained types of MWEs, high precisions above 90% have been achieved. This shows, enhanced with some linguistic filters, it can provide a practically useful tool for identifying and extracting MWEs. Furthermore, in our experiment, our tool demonstrated its capability of multilingual processing. With only minor adjustment, it can be ported to other languages. Meanwhile, further study is needed for a fuller understanding of the factors affecting the performance of statistical tools, including the text styles and topic/domains of the texts, etc.</Paragraph>
  </Section>
class="xml-element"></Paper>