<?xml version="1.0" standalone="yes"?>
<Paper uid="W06-0134">
  <Title>Sydney, July 2006. c(c)2006 Association for Computational Linguistics A Pragmatic Chinese Word Segmentation System</Title>
  <Section position="6" start_page="191" end_page="191" type="concl">
    <SectionTitle>
5 Conclusion
</SectionTitle>
    <Paragraph position="0"> We have briefly described our word segmentation system and NER system. We use word-based features in the whole processing. Our system has a good performance in terms of R iv measure, so this means that the trigram model with the smoothing algorithm can deal with the basic segmentation task well. However, the result in the bakeoff indicates that detecting out-of-vocabulary word seems to be a harder task than dealing with the segmentation-ambiguity task.</Paragraph>
    <Paragraph position="1"> The work in the future will concentrate on two sides: improving the NER performance and adding New Word Detection Algorithm.</Paragraph>
  </Section>
class="xml-element"></Paper>